A Blog by

Who’s the First Person in History Whose Name We Know?

Editor’s Note: This post has updated to clarify a sentence about the gender of the ancient writer.  

“It’s me!” they’d say, and they’d leave a sign. Leave it on the cave wall. Maybe as a prayer, maybe a graffito, we don’t know.

This was 30,000 years ago. Writing hadn’t been invented, so they couldn’t chalk their names on the rock. Instead, they’d flatten their hand, blow dust over it, and leave a silhouette like this:

a handprint is outlined in an orange/red pigment on the reproduction of the prototype fac simile of the cave Chauvet
Prototype fac simile of the cave Chauvet—Pont d’Arc, negative hand painted by blowing pigments. Photograph by Laurent CERINO, REA, Redux
Photograph by Laurent CERINO, REA, Redux

And for 30, 40 centuries across Europe, Asia, the Americas, and Australia, this is how cavemen, cavewomen, cave kids, hunters, nomads, farmers, and soldiers left their mark.

Picture of layers and layers of hands painted onto a cave wall in Argentina
Cave of the Hands, Patagonia, Province of Santa Cruz, Argentina. Photograph by
Javier Etcheverry, VWPics, Redux
Photograph by Javier Etcheverry, VWPics, Redux

Every one of these handprints belonged to an individual, presumably with a name, a history, and stories to tell. But without writing, we can’t know those stories. We call them hunter-gatherers, cave people, Neolithic tribes. We think of them in groups, never alone. Tens of thousands of generations come and go, and we can’t name a single person before 3200 B.C., not a one. Then, in Mesopotamia, writing appears, and after that people could record their words, sometimes in phonetic symbols so we could listen in, hear them talking and, for the first time, hear someone’s name—our first individual.

So who was it?

Who is the first person in the recorded history of the world whose name we know?

Just Guessing Here

Would it be a she or a he? (I’m figuring a he, because writing was a new thing, and males are usually the early adopters.) [*Please see note at bottom of post for more on this.]

Drawing of of man and a woman, the woman is crossed out.
All drawings by Robert Krulwich
Drawing by Robert Krulwich

Would he be a king? Warrior? Poet? Merchant? Commoner? (I’m guessing not a commoner. To be mentioned in an ancient document, he’d need a reputation, tools, and maybe a scribe. He wouldn’t be poor.)

Drawing of a king, a warrior, a poet, a merchant, and a commoner, with the commoner crossed out

Would he be a person of great accomplishment or just an ordinary Joe? (The odds favor a well-regarded person, someone who is mentioned often. Regular Joes, I figured, would pop up irregularly, while a great king, a leading poet, or a victorious general would get thousands of mentions.)

Drawing of a king sitting in a chair with a trident-like stick, looking at writing in front of him

So I trolled the internet, read some books, and to my great surprise—the first name in recorded history isn’t a king. Nor a warrior. Or a poet. He was, it turns out … an accountant. In his new book Sapiens: A Brief History of Humankind, Yuval Noah Harari goes back 33 centuries before Christ to a 5,000-year-old clay tablet found in Mesopotamia (modern Iraq). It has dots, brackets, and little drawings carved on it and appears to record a business deal.

Picture of an ancient tablet depicting beer production Inanna Temple in Uruk
MS1717, © The Schøyen Collection, Oslo and London http://www.schoyencollection.com/24-smaller-collections/wine-beer/ms-1717-beer-inanna-uruk
© The Schøyen Collection, Oslo and London

It’s a receipt for multiple shipments of barley. The tablet says, very simply:

29,086 measures barley 37 months Kushim

“The most probable reading of this sentence,” Harari writes, “is: ‘A total of 29,086 measures of barley were received over the course of 37 months. Signed, Kushim.’ ”

Drawing of a man facing the viewer with a speech bubble over his left shoulder that says " of “Oh, Kushim!”

So who was “Kushim”? The word might have been a job title, not a person (maybe kushim meant “barley assessor”) but check the video down below. It suggests that Kushim was indeed a guy, a record keeper who counted things for others—in short, an accountant. And if Kushim was his name, then with this tablet, Harari writes, “we are beginning to hear history through the ears of its protagonists. When Kushim’s neighbours called out to him, they might really have shouted, ‘Kushim!’”

It’s pretty clear Kushim was not famous, not hugely accomplished, certainly not a king. So all of my hunches were off.

But wait. The Kushim tablet is just one of tens of thousands of business records found on the deserts of Iraq. A single example is too random. We need more. So I keep looking and find what may be the second, third, and fourth oldest names we know of. They appear on a different Mesopotamian tablet.

Ancient stone tablet featuring a male figure, hunting dogs, and boars from Mesopotamia
Administrative tablet with cylinder seal impression of a male figure, hunting dogs, and boars. 3100-2900 B.C. Jamdat Nasr, Uruk III style, southern region, Mesopotamia. Clay, H. 2 in. (5.3 cm). Image copyright © The Metropolitan Museum of Art. Image source: Art Resource, NY
Image copyright © The Metropolitan Museum of Art. Image source: Art Resource, NY

Once again, they are not A-list ancients. Dated to around 3100 B.C.—about a generation or two after Kushim—the tablet’s heading is, “Two slaves held by Gal-Sal.” Gal-Sal is the owner. Next come the slaves, “En-pap X and Sukkalgir.” So now we’ve got four names: an accountant, a slave owner, and two slaves. No kings. They don’t show up for another generation or so.

Drawing of four individuals: an accountant, a slave owner, and two slaves

The predominance of ordinary Sumerians doesn’t surprise Harari. Five thousand years ago, most humans on Earth were farmers, herders, and artisans who needed to keep track of what they owned and what they owed—and that’s how writing started. It was a technology for regular people, not a megaphone for the powerful.

“It is telling,” Harari writes, “that the first recorded name in history belongs to an accountant, rather than a prophet, a poet, or a great conqueror.” Most of what people did back then was business.

Kings come, kings go, but keeping track of your barley—your sheep, your money, your property—that’s the real story of the world.


*Note from Robert Krulwich: I see that this column has offended a whole bunch of you. Yes, as many of you point out, my viewpoint was white, male (and hung up on fame and power) and many of you have serious, and totally legitimate arguments with my assumptions. Now that I read your comments, I’m a little surprised, and a touch ashamed of myself. But the thing is—those were my assumptions. They were wrong. I say so.

This is a blog. So it’s designed to be personal, and confessional. So I want you to know who’s talking to you, and if you think I’m way off base, by all means, let me know. And in the end, if you read the totality, my column and your responses, the story I wrote gets deeper and richer. You call me out on my assumptions, you offer some of your own, and what actually happened, what it was really like to be alive 5,300 years ago becomes… well, an argument among moderns about ancients that we will never meet.

Scholars aren’t unanimous about who’s name is oldest in the historical record. Yuval Noah Harari’s new book Sapiens: A Brief History of Humankind gives the crown to Kushim. The Oriental Institute at the University of Chicago goes for Gal-Sal and his slaves in their 2010-2011 annual report. Andrew Robinson, in his Writing and Script: A Very Short Introduction also champions Gal-Sal, but his book came earlier, so maybe Harari has scooped him. Here’s the video that argues for Kushim:

If the name Gal-Sal strikes some of you as familiar, it appears in the title of a 1942 Rita Hayworth/Victor Mature movie, My Gal Sal, about a songwriter who falls crazily in love with a singer on the vaudeville circuit named Sal (short for Sally Elliot). I watched it. It’s terrible. Kushim, meanwhile, survives. According to the blog Namespedia, it turns out that lots of Russian families call themselves Kushim to this day, and in the U.S., it’s a relatively popular first name. They’ve even got Kushim bar graphs!

A Blog by

Are Marmoset Monkeys Taking Turns To Talk?

When we talk to one another, we take turns. This simple rule seems to apply to all human conversation, whether the speakers are English city-dwellers or Namibian hunter-gatherers. One person speaks at a time and, barring the occasional interruption, we wait for our partner to finish before grabbing the conch. Timing is everything: cutting someone off is rude; leaving pregnant pauses is awkward. You need to leave a Goldilocks gap—something just right.

There are variations, certainly. New Yorkers are reputedly fond of “simultaneous speech” while Nordic cultures apparently love long, lingering pauses. But when Tanya Stivers analysed turn-taking across varied cultures, she found more similarities than differences. As I wrote in 2009:

“Stivers [collected] video recordings of conversations in ten different languages from five continents – from English to Korean, and from Tzeltal (a Mayan language spoken in Mexico) to Yeli-Dyne (a language of just 4,000 speakers used in Papua New Guinea). She found that… in all ten cultures, speakers shoot for as little silence as possible without speaking over each other, and the majority of answers follow questions after virtually no delay or overlap. The average delays certainly varied from language to language, but [the] extremes were only a quarter of a second off from the international average.”

The universal nature of turn-taking fascinated Asif Ghazanfar, a psychologist at Princeton University who studies monkey behaviour. “Taking turns acts as the foundation for more sophisticated forms of communication. You can’t share information if you’re constantly chattering over each other,” he says. “So how does that evolve?”

Our close relatives—the other great apes—provide few clues. They don’t actually vocalise very much and when they do, there’s no evidence that they take turns. So, Ghazanfar turned to another primate—the common marmoset, a tiny monkey that looks not unlike Back to the Future’s Doc Brown.

Although marmosets aren’t especially sophisticated communicators, they do regularly call to one another. Together with Daniel Takahashi and Darshana Narayanan, Ghazanfar placed 27 pairs of common marmosets  in opposite corners of a room, separated by an opaque curtain. Both monkeys called out, and although the pace of their exchanges was much slower than a human conversation, the team saw similarities in their rhythms.

For a start, they rarely interrupted one another. Each one waited for about 5 to 6 seconds after its partner finished before sounding off itself. The partners also ‘conversed’ with a steady rhythm, technically known as coupled oscillation. Both monkeys left a predictable interval between their calls, and their vocals slotted neatly into the silences created by their partner. And to confirm that they really are coordinated, the team showed that if one partner sped up or slowed down, the other followed suit.

“That’s what we do in conversation all the time,” says Ghazanfar. “If you speak to someone who’s speaking fast, you’ll start doing it too. We’re reporting the same for marmosets.”

Of course, there’s more to human turn-taking than that. We use sophisticated tricks to work out when it’s our turn to speak. We pay attention to grammar, meaning, inflection, body language and eye contact, and there’s no evidence that the marmosets are doing any of that. But nonetheless, the results are very similar—a coordinated vocal see-saw.

The marmosets also behaved in the same way whether they were paired with familiar cagemates  or complete strangers. That’s another feature they share with humans, and it sets them apart from, say, duetting birds, which only coordinate their vocals under very specific circumstances. “That’s not the case here,” says Ghazanfar. “One marmoset could have a conversation with any other marmoset.”

Common marmosets. Credit: Dario Sanchez.
Common marmosets. Credit: Dario Sanchez

But Margaret Wilson, a psychologist from the University of California, Santa Cruz who studies turn-taking in both humans and animals, is not convinced. “The paper doesn’t demonstrate turn-taking in any interesting sense,” she says.  “I think the authors have failed to appreciate just how weird human turn-taking is.” Wilson explains the weirdness beautifully, so I’m going to yield the floor without interruptions:

When humans take turns, there is a cyclic structure to the extremely short gaps between speakers’ utterances.  A between-turn gap of, say, 200 milliseconds is more likely to be broken by the second speaker at certain regular intervals (say, odd multiples of 50 ms) than during the “troughs” between those intervals.  That is, short silences are not of arbitrary length, but reflect a cyclic passing back and forth of who has the “right” to speak next.  The troughs represent moments when the right to speak has shifted back to the original speaker, hence the second speaker inhibits speech during those fractions of a second.  And this is happening at the order of tens of milliseconds.  This “structured silence” can only be explained by extremely tight coupling of some oscillatory mechanism in the brains of the two speakers.“

And there’s no hint of that complexity in the marmosets’ exchanges. The cycles in their conversations span their actual calls rather than just the gaps between them. That just means their timing’s not completely random. And with long silences between turns, they could just have been calling and then responding to their partner’s call, as many other animals do.

But if there’s one thing Wilson agrees with, it’s that the question’s worth asking. “Turn-taking is fundamental to human conversation, so the question of whether it occurs in other social animals is extremely interesting,” she says.

Consider that humans talk so much more than other apes. Many scientists have suggested that this vocal sophistication is rooted in manual gestures—the arm and hand movements that chimps and gorillas use a lot. These became increasingly complex and eventually, the brain circuits for gestures got glommed onto vocals. But Ghazanfar isn’t a fan of this idea. “They have to come up with some magical thing that switches from manual to speech,” he says.

If marmosets take turns, that points to a different hypothesis. Like humans, they are cooperative breeders. Males and females work together to raise their young, typically as a monogamous pair and often with help from older siblings. “The idea is that this strategy of cooperative breeding specifically makes them more friendly,” says Ghazanfar. Turn-taking may be a symptom of this temperament—a by-product of breeding habits that make for a generally more cooperative primate.

Maybe the same thing happened during human evolution? “There could have been a tweak in the way we raise our offspring, which led to more prosocial behaviour,” says Ghazanfar. “And once you have that general prosociality, you may be more inclined to make more contact with other members of the species.”

It’s an interesting scenario, and one that has parallels in domestic dogs. There’s a popular idea that dogs evolved from wolves that were drawn to human settlements, perhaps to scavenge off our garbage. Individauly with more docile temperaments were best-suited to these forays, and gradually became better at reading human cues and gestures. You select for a certain temperament, and many other traits get yanked along for the evolutionary ride.

Of course, this is still conjecture, and the common marmosets are just one (contested) data point. The next step would be to check for turn-taking in other cooperatively breeding primates, such as tamarins, or related species that don’t share the same reproductive habits.

Reference: Takahashi, Narayanan & Ghazanfar. 2013. Coupled Oscillator Dynamics of Vocal Turn-Taking in Monkeys. Current Biology http://dx.doi.org/10.1016/j.cub.2013.09.005

A Blog by

How Forensic Linguistics Outed J.K. Rowling (Not to Mention James Madison, Barack Obama, and the Rest of Us)

Earlier this week, the UK’s Sunday Times rocked the publishing world by revealing that Robert Galbraith, the first-time author of a new crime novel called The Cuckoo’s Calling, is none other than J.K. Rowling, the superstar author of the Harry Potter series. Then the New York Times told the story of how the Sunday Times’s arts editor, Richard Brooks, had figured it out.

One of Brooks’s colleagues got an anonymous tip on Twitter claiming that Galbraith was Rowling. The tipster’s Twitter account was then swiftly deleted. Before confronting the publisher with the question, Brooks’s team did some web sleuthing. They found that the two authors shared the same publisher and agent. And, after consulting with two computer scientists, they discovered that The Cuckoo’s Calling and Rowling’s other books show striking linguistic similarities. Satisfied that the Twitter tipster was right, Brooks reached out to Rowling. Finally, on Saturday morning, as the New York Times reports, “he received a response from a Rowling spokeswoman, who said that she had ‘decided to fess up’.”

While the literary world was buzzing about whether that anonymous tipster was actually Rowling’s publisher, Little, Brown and Company (it wasn’t), I wanted to know how those computer scientists did their mysterious linguistic analyses. I called both of them yesterday and learned not only how the Rowling investigation worked, but about the fascinating world of forensic linguistics.

With computers and sophisticated statistical analyses, researchers are mining all sorts of famous texts for clues about their authors. Perhaps more surprising: They’re are also mining not-so-famous texts, like blogs, tweets, Facebook updates and even Amazon reviews for clues about people’s lifestyles and buying habits. The whole idea is so amusingly ironic, isn’t it? Writers choose words deliberately, to convey specific messages. But those same words, it turns out, carry personal information that we don’t realize we’re giving out.

“There’s a kind of fascination with the thought that a computer sleuth can discover things that are hidden there in the text. Things about the style of the writing that the reader can’t detect and the author can’t do anything about, a kind of signature or DNA or fingerprint of the way they write,” says Peter Millican of Oxford University, one of the experts consulted by the Sunday Times.

Cal Flyn, a reporter with the Sunday Times, sent email requests to Millican and to Patrick Juola, a computer scientist at Duquesne University in Pittsburgh. Flyn told them the hypothesis — that Galbraith was Rowling — and gave them the text of five books to test that hypothesis. Those books included Cuckoo, obviously, as well as a novel by Rowling called The Casual Vacancy. The other three were all, like Cuckoo, British crime novels: The St. Zita Society by Ruth Rendell, The Private Patient by P.D. James, and The Wire in the Blood by Val McDermid.

Juola ran each book (or, more precisely, the sequence of tens of thousands of words that make up a book) through a computer program that he and his students have been working on for more than 10 years, dubbed JGAAP. He compared Cuckoo to the other books using four different analyses, each focused on a different aspect of writing.

One of those tests, for example, compared all of the word pairings, or sets of adjacent words, in each book. “That’s better than individual words in a lot of ways because it captures not just what you’re talking about but also how you’re talking about it,” Juola says. This test could show, for example, the types of things an author describes as expensive: an expensive car, expensive clothes, expensive food, and so on. “It might be that this is a word that everyone uses, like expensive, but depending on what you’re focusing on, it [conveys] a different idea.”

Juola also ran a test that searched for “character n-grams”, or sequences of adjacent characters. He focused on 4-grams, or four-letter sequences. For example, a search for the sequence “jump” would bring up not only jump, but jumps, jumped, and jumping. “That lets us look at concepts and related words without worrying about tense and conjugation,” he says.

Those two tests turn up relatively rare words. But even a book’s most common words — words like a, and, of, the — leave a hidden signature. So Juola’s program also tallied the 100 most common words in each book and compared the small differences in frequency. One book might have used the word “the” six percent of the time, while another uses it only 4 percent.

Juola’s final test completely separates a word from its meaning, by sorting words simply by their length. What fraction of a book is made of three-letter words, or eight-letter words? These distributions are fairly similar from book to book, but statistical analyses can dig into the subtle differences. And this particular test “was very characteristically Rowling,” Juola says. “Word lengths was one of the strongest pieces of evidence that [Cuckoo] was Rowling.”

It took Juola about an hour and a half to do all of these word-crunchings, and all four tests suggested that Cuckoo was more similar to Rowling’s Casual Vacancy than the other books. And that’s what he relayed back to Flyn. Still, though, he wasn’t totally confident in the result. After all, he had no way of knowing whether the real author was somebody who wasn’t in the comparison set of books who happened to write like Rowling does. “It could have been somebody who looked like her. That’s the risk with any police line-up, too,” he says.

Meanwhile, across the pond, Peter Millican was running a parallel Rowling investigation. After getting Flyn’s email, Millican told her he needed more comparison data, so he ended up with an additional book from each of the four known authors (using Harry Potter and the Deathly Hallows as the second known Rowling book). He ran those eight books, plus Cuckoo, into his own linguistics software program, called Signature.

Signature includes a fancy statistical method called principal component analysis to compare all of the books on six features: word length, sentence length, paragraph length, letter frequency, punctuation frequency, and word usage.

Word frequency tests can be done in different ways. Juola, as I described, looked at word pairings and at the most common words. Another approach that can be quite definitive, Millican says, is a comparison of rare words. The classical example concerns the Federalist Papers, a series of essays written by Alexander Hamilton, James Madison, and John Jay during the creation of the U.S. Constitution. In 1963, researchers used word counts to determine the authorship of 12 of these essays that were written by either Madison or Hamilton. They found that Madison’s essays tended to use “whilst” and never “while”, and “on” rather than “upon”. Hamilton, in contrast, tended to use “while”, not “whilst”, and used “on” and “upon” at the same frequency. The 12 anonymous papers never used “while” and rarely used “upon”, pointing strongly to Madison as the author.

Millican found a few potentially distinctive words in his Rowling investigation. The other authors tended to use the words “course” (as in, of course), “someone” and “realized” a bit more than Rowling did. But the difference wasn’t statistically significant enough for Millican to run with it. So, like Juola, he turned to the most common words. Millican pulled out the 500 most common words in each book, and then went through and manually removed the words that were subject-specific, such as “Harry”, “wand”, and “police”.

Of all of the tests he can run with his program, Millican finds these word usage comparisons most compelling. “You end up with a graph, and on the graph it’s absolutely clear that Cuckoo’s Calling is lining up with Harry Potter. And it’s also clear that the Ruth Rendell books are close together, the Val McDermid books are close together, and so on,” he says. “It is identifying something objective that’s there. You can’t easily describe in English what it’s detecting, but it’s clearly detecting a similarity.”

On all of Millican’s tests, Cuckoo turned out to be most similar to a known Rowling book, and on four of them, both Rowling books were closer than any of the others. Millican got the files around 8pm on Friday night. Five hours later, he emailed the Sunday Times. “I said, ‘I’m pretty certain that if it’s one of these four authors, it’s Rowling.'”

This isn’t the first time that Millican has found himself in the middle of a high-profile authorship dispute. In the fall of 2008, just a couple of weeks before the U.S. presidential election, he got an email from the brother-in-law of a Republican congressman from Utah. He told him that they had used his Signature software (which is downloadable from his website) to show that Barack Obama’s book, Dreams from my Father, could have been written by Bill Ayers, a domestic terrorist. “They were planning to have a press conference in Washington to expose Obama one week before the election and got in touch with me,” Millican recalls, chuckling. “It was quite a strange situation to be in.”

Millican re-ran the analysis and definitively showed that Dreams was not, in fact, written by Ayers (you can read more about what he did here).

Juola told me some crazy stories, too. He once worked on a legal case in which a man had written a set of anonymous newspaper articles that were critical of a foreign government. He was facing deportation proceedings in the United States, and knew that if he was deported then the secret police in said foreign government would be waiting for him at the airport. Juola’s analyses confirmed that the anonymous articles were, in fact, written by the man. And because of that, he was permitted to stay in the U.S. “We were able to establish his identity to the satisfaction of the judge,” Juola says.

That story, he adds, shows how powerful this kind of science can be. “There are a lot of real controversies with real consequences for the people involved that are a lot more important than just, did this obscure novel get written by this particular famous author?”

The words of many of us, in fact, are probably being mined at this very moment. Some researchers, Juola told me, are working on analyzing product reviews left on websites like Amazon.com. These investigations could root out phony glowing reviews left by company representatives, for example, or reveal valuable demographic patterns.

“They might say, hmmm, that’s funny, it looks like all of the women from the American West are rating our product a star and a half lower than men from the northeast, so obviously we need ot do some adjustment of our advertisements,” he says. “Not many companies are going to admit to doing this kind of thing, but anytime you’ve got some sort of investigation going on, whether police or security clarance or a job application, one of the things you’re going to look at is somebody’s public profile on the web. Anything is fair game.”

In fact, it was a good thing the original tipster of the Rowling news deleted his or her Twitter account, Juola says. “If we still had the account, we could have looked at the phrasings to see if it corresponded to anyone who works at the publishing house.”

A Blog by

British family’s problems hint at a gene involved in linking language and meaning

We’ve all had that annoying feeling when we fail to find a word that’s just at the tip of our tongues. Usually, these moments are passing nuisances, but they are a more severe impediment for a British family known as JR. Eight of them suffer from an unusual problem with “semantic cognition” – the ability to bind words to their meanings during thought or communication.

They can’t remember words, names, or topics of conversation – all of us get this, but the JR family experiences a more extreme version. They make errors in everyday conversations when they use words with related meanings in the wrong places. Their comprehension falters to the extent that reading books or following films is hard work.

These difficulties have caused them much social anxiety, and hampered their ability to cope with school and work. But for scientists, they are undeniably exciting because they seem to stem from a single errant gene. If that’s the case, the gene apparently affects the intertwining of concepts and language, but not any other mental abilities – the affected family members are otherwise intelligent and articulate. The JR family could lead us to new insights about language, thought and memory, just as similar families have done in the past.


A Blog by

The power of nouns – tiny word change increases voter turnout

Countries around the world have tried many tactics to encourage people to vote, from easier access to polling stations to mandatory registration. But Christopher Bryan from Stanford University has found a startlingly simple weapon for increasing voter turnout – the noun. Through a simple linguistic tweak, he managed to increase the proportion of voters in two groups of Americans by at least 10 percentage points.

During the 2008 presidential election, Bryan recruited 34 Californians who were eligible to vote but hadn’t registered yet. They all completed a survey which, among other questions, asked them either “How important is it to you to be a voter in the upcoming election?” or “How important is it to you to vote in the upcoming election?”

It was the tiniest of tweaks – the noun-focused “voter” versus the verb-focused “vote” – but it was a significant one. Around 88% of the noun group said they were very or extremely interested in registering to vote, compared to just 56% of the verb group.


A Blog by

The cultural genome: Google Books reveals traces of fame, censorship and changing languages

Four_percent_of_all_the_booJust as petrified fossils tell us about the evolution of life on earth, the words written in books narrate the history of humanity. They words tell a story, not just through the sentences they form, but in how often they occur. Uncovering those tales isn’t easy – you’d need to convert books into a digital format so that their text can be analysed and compared. And you’d need to do that for millions of books.

Fortunately, that’s exactly what Google have been doing since 2004. Together with over 40 university libraries, the internet titan has thus far scanned over 15 million books, creating a massive electronic library that represents 12% of all the books ever published. All the while, a team from Harvard University, led by Jean-Baptiste Michel and Erez Lieberman Aiden have been analysing the flood of data.

Their first report is available today. Although it barely scratches the surface, it’s already a tantalising glimpse into the power of the Google Books corpus. It’s a record of human culture, spanning six centuries and seven languages. It shows vocabularies expanding and grammar evolving. It contains stories about our adoption of technology, our quest for fame, and our battle for equality. And it hides the traces of tragedy, including traces of political suppression, records of past plagues, and a fading connection with our own history.


A Blog by

New languages evolve in rapid bursts


This is an old article, reposted from the original WordPress incarnation of Not Exactly Rocket Science. I’m on holiday for the moment, but you can expect a few new pieces here and there (as well as some exciting news…)

The birth of new languages is accompanied by a burst of rapid evolution consisting of large changes in vocabulary that are followed by long periods of relatively slower change.

Languages are often compared to living species because of the way in which they diverge into new tongues over time in an ever-growing linguistic tree. Some critics have claimed that this comparison is a superficial one, a nice metaphor but nothing more.

But the new study by Quentin Atkinson, now at the University of Oxford, suggests that languages evolve at a similar stop-and-start pace, which uncannily echoes a long-standing theory in biology, known as ‘punctuated equilibrium’. The theory’s followers claim that life on Earth also evolved at an uneven pace, full of rapid bursts and slow periods.


A Blog by

New Nicaraguan sign language shows how language affects thought

NicaraguaOne of the signs for “Nicaragua”. Photo by Ann Serghas

In the 1970s, a group of deaf Nicaraguan schoolchildren invented a new language. The kids were the first to enrol in Nicaragua’s new wave of special education schools. At first, they struggled with the schools’ focus on Spanish and lip-reading, but they found companionship in each other. It was the first time that deaf people from all over the country could gather in large numbers and through their interactions – in the schoolyard and the bus – Nicaraguan Sign Language (NSL) spontaneously came into being.

NSL is not a direct translation of Spanish – it is a language in its own right, complete with its own grammar and vocabulary. Its child inventors created it naturally by combining and adding to gestures that they had used at home. Gradually, the language became more regular, more complex and faster. Ever since, NSL has been a goldmine for scientists, providing an unparalleled opportunity to study the emergence of a new language. And in a new study led by Jennie Pyers from Wellesley College, it even tells us how language shapes our thought.

By studying children who learned NSL at various stages of its development, Pyers has shown that the vocabulary they pick up affects the way they think. Specifically, those who learned NSL before it developed specific gestures for left and right perform more poorly on a spatial awareness test than children who grew up knowing how to sign those terms.


A Blog by

Requests work better than orders, even when we’re asking or ordering ourselves

We like to be in control of our own lives, and some of us have an automatic rebellious streak when we’re told what to do. We’re less likely to do a task if we’re ordered to do it than if we make the choice of our own volition. It seems that this effect is so strong that it even happens when the people giving the orders are… us.

In a set of three experiments, Ibrahim Senay from the University of Illinois has shown that people do better at a simple task if ask themselves whether they’ll do it than if they simply tell themselves to do so. Even a simple reversal of words – “Will I” compared to “I will” – can boost motivation and performance.

Therapists and managers alike are taught to ask people open questions that prompt them to think about problems for themselves, rather than having solutions imposed upon them. Senay’s work suggests that this approach would work even if we’re counselling or managing ourselves. When we question ourselves about our deeds and choices, we’re more likely to consider our motivations for doing something and feel like we’re in control of our actions. The effect is small but significant.


A Blog by

Boom-boom-krak-oo – Campbell’s monkeys combine just six ‘words’ into rich vocabulary

Many human languages achieve great diversity by combining basic words into compound ones – German is a classic example of this. We’re not the only species that does this. Campbell’s monkeys have just six basic types of calls but they have combined them into one of the richest and most sophisticated of animal vocabularies.

By chaining calls together in ways that drastically alter their meaning, they can communicate to each other about other falling trees, rival groups, harmless animals and potential threats. They can signal the presence of an unspecified threat, a leopard or an eagle, and even how imminent the danger is. It’s a front-runner for the most complex example of animal “proto-grammar” so far discovered.

Many studies have shown that the chirps and shrieks of monkeys are rich in information, ever since Dorothy Cheney and Robert Seyfarth’s seminal research on vervet monkeys. They showed that vervets have specific calls for different predators – eagles, leopards and snakes – and they’ll take specific evasive manoeuvres when they hear each alarm.

Campbell’s monkeys have been equally well-studied. Scientists used to think that they made two basic calls – booms and hacks – and that the latter were predator alarms. Others then discovered that the order of the calls matters, so adding a boom before a hack cancels out the predator message. It also turned out that there were five distinct types of hack, including some that were modified with an -oo suffix. So Campbell’s monkeys not only have a wider repertoire of calls than previously thought, but they can also combine them in meaningful ways.

Now, we know that the males make six different types of calls, comically described as boom (B), krak (K), krak-oo (K+), hok (H), hok-oo (H+) and wak-oo (W+). To decipher their meaning,  Karim Ouattara spent 20 months in the Ivory Coast’s Tai National Park studying the wild Campbell’s monkeys from six different groups. Each consists of a single adult male together with several females and youngsters. And it’s the males he focused on.

With no danger in sight, males make three call sequences. The first – a pair of booms – is made when the monkey is far away from the group and can’t see them. It’s a summons that draws the rest of the group towards him. Adding a krak-oo to the end of the boom pair changes its meaning. Rather than “Come here”, the signal now means “Watch out for that branch”. Whenever the males cried “Boom-boom-krak-oo”, other monkeys knew that there were falling trees or branches around (or fighting monkeys overhead that could easily lead to falling vegetation). 

Interspersing the booms and krak-oos with some hok-oos changes the meaning yet again. This call means “Prepare for battle”, and it’s used when rival groups or strange males have showed up. In line with this translation, the hok-oo calls are used far more often towards the edge of the monkeys’ territories than they are in the centre. The most important thing about this is that hok-oo is essentially meaningless. The monkeys never say it in isolation – they only use it to change the meaning of another call.

But the most complex calls are reserved for threats. When males know that danger is afoot but don’t have a visual sighting (usually because they’ve heard a suspicious growl or an alarm from other monkeys), they make a few krak-oos. 

If they know it’s a crowned eagle that endangers the group, they combine krak-oo and wak-oo calls. And if they can actually see the bird, they add hoks and hok-oos into the mix – these extra components tell other monkeys that the peril is real and very urgent.  Leopard alarms were always composed of kraks, and sometimes krak-oos. Here, it’s the proportion of kraks that signals the imminence of danger – the males don’t make any if they’ve just heard leopard noises, but they krak away if they actually see the cat. 

The most important part of these results is the fact that calls are ordered in very specific ways. So boom-boom-krak-oo means a falling branch, but boom-krak-oo-boom means nothing. Some sequences act as units that can be chained together to more complicated ones – just as humans use words, clauses and sentences. They can change meaning by adding meaningless calls onto meaningful ones (BBK+ for falling wood but BBK+H+ for neighbours) or by chaining meaningful sequences together (K+K+ means leopard but W+K+ means eagle).

It’s tempting to think that monkeys have hidden linguistic depths to rival those of humans but as Ouattara says, “This system pales in contrast to the communicative power of grammar.” They monkeys’ repertoire may be rich, but it’s still relatively limited and they don’t take full advantage of their vocabulary. They can create new meanings by chaining calls together, but never by inverting their order (e.g. KB rather than BK).  Our language is also symbolic. I can tell you about monkeys even though none are currently scampering about my living room, but Ouattara only found that Campbell’s monkeys “talk” about things that they actually see.

Nonetheless, you have to start somewhere, and the complexities of human syntax probably have their evolutionary origins in these sorts of call combinations. So far, the vocabulary of Campbell’s monkeys far outstrips those of other species, but this may simply reflect differences in research efforts. Other studies have started to find complex vocabularies in other forest-dwellers like Diana monkeys and putty-nosed monkeys. Ouattara thinks that forest life, with many predators and low visibility, may have provided strong evolutionary pressures for monkeys to develop particularly sophisticated vocal skills.

And there are probably hidden depths to the sequences of monkey calls that we haven’t even begun to peer into yet. For instance, what calls do female Campbell’s monkeys make? Even for the males, the meanings in this study only become apparent after months of intensive field work and detailed statistical analysis. The variations that happen on a call-by-call basis still remain a mystery to us. The effect would be like looking at Jane Austen’s oeuvre and concluding, “It appears that these sentences signify the presence of posh people”.

Reference: PNAS doi:10.1073/pnas.0908118106

More on monkey business (clearly, I need more headline variation):


A Blog by

Revisiting FOXP2 and the origins of language

Today, a new paper published in Nature adds another chapter to the story of FOXP2, a gene with important roles in speech and language. The FOXP2 story is a fascinating tale that I covered in New Scientist last year. It’s one of the pieces I’m proudest of so I’m reprinting it here with kind permission from Roger Highfield, and with edits incorporating new discoveries since the time of writing.

The FOXP2 Story (2009 edition)  

Imagine an orchestra full of eager musicians which, thanks to an incompetent conductor, produces nothing more than an unrelieved cacophony. You’re starting to appreciate the problem faced by a British family known as KE. About half of its members have severe difficulties with language. They have trouble with grammar, writing and comprehension, but above all they find it hard to coordinate the complex sequences of face and mouth movements necessary for fluid speech.

Thanks to a single genetic mutation, the conductor cannot conduct, and the result is linguistic chaos. In 2001, geneticists looking for the root of the problem tracked it down to a mutation in a gene they named FOXP2. Normally, FOXP2 coordinates the expression of other genes, but in affected members of the KE family, it was broken.

It had long been suspected that language has some basis in genetics, but this was the first time that a specific gene had been implicated in a speech and language disorder. Overeager journalists quickly dubbed FOXP2 “the language gene” or the “grammar gene”. Noting that complex language is a characteristically human trait, some even speculated that FOXP2 might account for our unique position in the animal kingdom. Scientists were less gushing but equally excited – the discovery sparked a frenzy of research aiming to uncover the gene’s role.

Several years on, and it is clear that talk of a “language gene” was premature and simplistic. Nevertheless, FOXP2 tells an intriguing story. “When we were first looking for the gene, people were saying that it would be specific to humans since it was involved in language,” recalls Simon Fisher at the University of Oxford, who was part of the team that identified FOXP2 in the KE family. In fact, the gene evolved before the dinosaurs and is still found in many animals today: species from birds to bats to bees have their own versions, many of which are remarkably similar to ours. “It gives us a really important lesson,” says Fisher. “Speech and language didn’t just pop up out of nowhere. They’re built on very highly conserved and evolutionarily ancient pathways.”

Two amino acids, two hundred thousand years

The first team to compare FOXP2 in different species was led by Wolfgang Enard from the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. In 2001, they looked at the protein that FOXP2 codes for, called FOXP2, and found that our version differs from those of chimpanzees, gorillas and rhesus macaques by two amino acids out of a total of 715, and from that of mice by three. This means that the human version of FOXP2 evolved recently and rapidly: only one amino acid changed in the 130 million years since the mouse lineage split from that of primates, but we have picked up two further differences since we diverged from chimps, and this seems to have happened only with the evolution of our own species at most 200,000 years ago.

The similarity between the human protein FOXP2 and that of other mammals puts it among the top 5 per cent of the most conserved of all our proteins. What’s more, different human populations show virtually no variation in their FOXP2 gene sequences. Last year, Enard’s colleague Svante Pääbo made the discovery that Neanderthals also had an identical gene, prompting questions over their linguistic abilities (see “Neanderthal echoes below).

“People sometimes think that the mutated FOXP2 in the KE family is a throwback to the chimpanzee version, but that’s not the case,” says Fisher. The KEs have the characteristically human form of the gene. Their mutation affects a part of the FOXP2 protein that interacts with DNA, which explains why it has trouble orchestrating the activity of other genes.

There must have been some evolutionary advantage associated with the human form of FOXP2, otherwise the two mutations would not have spread so quickly and comprehensively through the population. What this advantage was, and how it may have related to the rise of language, is more difficult to say. Nevertheless, clues are starting to emerge as we get a better picture of what FOXP2 does – not just in humans but in other animals too.

During development, the gene is expressed in the lungs, oesophagus and heart, but what interests language researchers is its role in the brain. Here there is remarkable similarity across species: from humans to finches to crocodiles, FOXP2 is active in the same regions. With no shortage of animal models to work with, several teams have chosen songbirds due to the similarities between their songs and human language: both build complex sequences from basic components such as syllables and riffs, and both forms of vocalisation are learned through imitation and practice during critical windows of development.

Babbling birds

All bird species have very similar versions of FOXP2. In the zebra finch, its protein is 98 per cent identical to ours, differing by just eight amino acids. It is particularly active in a part of the basal ganglia dubbed “area X”, which is involved in song learning. Constance Scharff at the Max Planck Institute for Molecular Genetics in Berlin, Germany, reported that finches’ levels of FOXP2 expression in area X are highest during early life, which is when most of their song learning takes place. In canaries, which learn songs throughout their lives, levels of the protein shoot up annually and peak during the late summer months, which happens to be when they remodel their songs.

So what would happen to a bird’s songs if levels of the FOXP2 protein in its area X were to plummet during a crucial learning window? Scharff found out by injecting young finches with a tailored piece of RNA that inhibited the expression of the FOXP2 gene. The birds had difficulties in developing new tunes and their songs became garbled: they contained the same component “syllables” as the tunes of their tutors, but with syllables rearranged, left out, repeated incorrectly or sung at the wrong pitch.

The cacophony produced by these finches bears uncanny similarities to the distorted speech of the afflicted KE family members, making it tempting to pigeonhole FOXP2 as a vocal learning gene – influencing the ability to learn new communication sounds by imitating others. But that is no more accurate than calling it a “language gene”. For a start, songbird FOXP2 has no characteristic differences to the gene in non-songbirds. What’s more, among other species that show vocal learning, such as whales, dolphins and elephants, there are no characteristic patterns of mutation in their FOXP2 that they all share.

Instead, consensus is emerging that FOXP2 probably plays a more fundamental role in the brain. Its presence in the basal ganglia and cerebellums of different animals provides a clue as to what that role might be. Both regions help to produce precise sequences of muscle movements. Not only that, they are also able to integrate information coming in from the senses with motor commands sent from other parts of the brain. Such basic sensory-motor coordination would be vital for both birdsong and human speech. So could this be the key to understanding FOXP2?

Moving mice

New work by Fisher and his colleagues supports this idea. In 2008, his team engineered mice to carry the same FOXP2 mutation that affects the KE family, rendering the protein useless. Mice with two copies of the dysfunctional FOXP2 had shortened lives, characterised by motor disorders, growth problems and small cerebellums. Mice with one normal copy of FOXP2 and one faulty copy (as is the case in the affected members of the KE family) seemed outwardly healthy and capable of vocalisation, but had subtle defects.

For example, they found it difficult to acquire new motor skills such as learning to run faster on a tilted running wheel. An examination of their brains revealed the problem. The synapses connecting neurons within the cerebellum, and those in a part of the basal ganglia called the striatum in particular, were severely flawed. The signals that crossed these synapses failed to develop the long-term changes that are crucial for memory and learning. The opposite happened when the team engineered mice to produce a version of FOXP2 with the two characteristically human mutations. Their basal ganglia had neurons with longer outgrowths (dendrites) that were better able to strengthen or weaken the connections between them.

A battery of over 300 physical and mental tests showed that the altered mice were generally healthy. While they couldn’t speak like their cartoon equals, their central nervous system developed in different ways, and they showed changes in parts of the brain where FOXP2 is usually expressed (switched on) in humans.

Their squeaks were also subtly transformed. When mouse babies are moved away from their nest, they make ultrasonic distress calls that are too high for us to hear, but that their mothers pick up loudly and clearly. The altered Foxp2 gene subtly changed the structure of these alarm calls. We won’t know what this means until we get a better understanding of the similarities between mouse calls and human speech.

For now, the two groups of engineered mice tentatively support the idea that human-specific changes to FOXP2 affect aspects of speech, and strongly support the idea that they affect aspects of learning. “This shows, for the first time, that the [human-specific] amino-acid changes do indeed have functional effects, and that they are particularly relevant to the brain,” explains Fisher. “FOXP2 may have some deeply conserved role in neural circuits involved in learning and producing complex patterns of movement.” He suspects that mutant versions of FOXP2 disrupt these circuits and cause different problems in different species.

Pääbo agrees. “Language defects may be where problems with motor coordination show up most clearly in humans, since articulation is the most complex set of movements we make in our daily life,” he says. These circuits could underpin the origins of human speech, creating a biological platform for the evolution of both vocal learning in animals and spoken language in humans.

Holy diversity, Batman

The link between FOXP2 and sensory-motor coordination is bolstered further by research in bats. Sequencing the gene in 13 species of bats, Shuyi Zhang and colleagues from the East China Normal University in Shanghai discovered that it shows incredible diversity. Why would bats have such variable forms of FOXP2 when it is normally so unwavering in other species?

Zhang suspects that the answer lies in echolocation. He notes that the different versions seem to correspond with different systems of sonar navigation used by the various bat species. Although other mammals that use echolocation, such as whales and dolphins, do not have special versions of FOXP2, he points out that since they emit their sonar through their foreheads, these navigation systems have fewer moving parts. Furthermore, they need far less sensory-motor coordination than flying bats, which vocalise their ultrasonic pulses and adjust their flight every few milliseconds, based on their interpretation of the echoes they receive.

These bats suggest that FOXP2 is no more specific to basic communication than it is to language, and findings from other species tell a similar tale. Nevertheless, the discovery that this is an ancient gene that has assumed a variety of roles does nothing to diminish the importance of its latest incarnation in humans.

Since its discovery, no other gene has been convincingly implicated in overt language disorders. FOXP2 remains our only solid lead into the genetics of language. “It’s a molecular window into those kinds of pathways – but just one of a whole range of different genes that might be involved,” says Fisher. “It’s a starting point for us, but it’s not the whole story.” He has already used FOXP2 to hunt down other key players in language.

The executive’s minions

FOXP2 is a transcription factor, which activates some genes while suppressing others. Identifying its targets, particularly in the human brain, is the next obvious step. Working with Daniel Geschwind at the University of California, Los Angeles, Fisher has been trying to do just that, and their preliminary results indicate just what a massive job lies ahead. On their first foray alone, the team looked at about 5000 different genes and found that FOXP2 potentially regulates hundreds of these.

Some of these target genes control brain development in embryos and its continuing function in adults. Some affect the structural pattern of the developing brain and the growth of neurons. Others are involved in chemical signalling and the long-term changes in neural connections that enable to learning and adaptive behaviour. Some of the targets are of particular interest, including 47 genes that are expressed differently in human and chimpanzee brains, and a slightly overlapping set of 14 targets that have evolved particularly rapidly in humans.

Most intriguingly, Fisher says, “we have evidence that some FOXP2 targets are also implicated in language impairment.” Last year, Sonja Vernes in his group showed that FOXP2 switches off CNTNAP2, a gene involved in not one but two language disorders – specific language impairment (SLI) and autism. Both affect children, and both involve difficulties in picking up spoken language skills. The protein encoded by CNTNAP2 is deployed by nerve cells in the developing brain. It affects the connections between these cells and is particularly abundant in neural circuits that are involved in language.

Verne’s discovery is a sign that the true promise of FOXP2’s discovery is being fulfilled – the gene itself has been overly hyped, but its true worth lies in opening a door for more research into genes involved in language. It was the valuable clue that threw the case wide open. CNTNAP2 may be the first language disorder culprit revealed through FOXP2 and it’s unlikely to be the last.

Most recently, Dan Geschwind compared the network of genes that are targeted by FOXP2 in both chimps and humans. He found that the two human-specific amino acids within this executive protein have radically altered the set of genetic minions that it controls.

The genes that are directed by human FOXP2 are a varied cast of players that influence the development of the head and face, parts of the brain involved in motor skills, the growth of cartilage and connective tissues, and the development of the nervous system. All those roles fit with the idea that our version of FOXP2 has been a lynchpin in evolving the neural circuits and physical structures that are important for speech and language.

The FOXP2 story is far from complete, and every new discovery raises fresh questions just as it answers old ones. Already, this gene has already taught us important lessons about evolution and our place in the natural world. It shows that our much vaunted linguistic skills are more the result of genetic redeployment than out-and-out innovation. It seems that a quest to understand how we stand apart from other animals is instead leading to a deeper appreciation of what unites us.

Box – Neanderthal echoes

The unique human version of the FOXP2 gives us a surprising link with one extinct species. Last year, Svante Pääbo’s group at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, extracted DNA from the bones of two Neanderthals, one of the first instances of geneticists exploring ancient skeletons for specific genes. They found that Neanderthal FOXP2 carries the same two mutations as those carried by us – mutations accrued since our lineage split from chimps between 6 and 5 million years ago.

Pääbo admits that he “struggled” to interpret the finding: the Neanderthal DNA suggests that the modern human’s version of FOXP2 arose much earlier than previously thought. Comparisons of gene sequences of modern humans with other living species had put the origins of human FOXP2 between 200,000 and 100,000 years ago, which matches archaeological estimates for the emergence of spoken language. However, Neanderthals split with humans around 400,000 years ago, so the discovery that they share our version of FOXP2 pushes the date of its emergence back at least that far.

“We believe there were two things that happened in the evolution of human FOXP2,” says Pääbo. “The two amino acid changes – which happened before the Neanderthal-human split – and some other change which we don’t know about that caused the selective sweep more recently.” In other words, the characteristic mutations that we see in human FOXP2 may indeed be more ancient than expected, but the mutated gene only became widespread and uniform later in human history. While many have interpreted Pääbo’s findings as evidence that Neanderthals could talk, he is more cautious. “There’s no reason to assume that they weren’t capable of spoken language, but there must be many other genes involved in speech that we yet don’t know about in Neanderthals.”


A Blog by

Native language shapes the melody of a newborn baby’s cry

Telling the difference between a German and French speaker isn’t difficult. But you may be more surprised to know that you could have a good stab at distinguishing between German and French babies based on their cries. The bawls of French newborns tend to have a rising melody, with higher frequencies becoming more prominent as the cry progresses. German newborns tend to cry with a falling melody.

These differences are apparent just three days out of the womb. This suggests that they pick up elements of their parents’ language before they’re even born, and certainly before they start to babble themselves.

Birgit Mampe from the University of Wurzburg analysed the cries of 30 French newborns and 30 German ones, all born to monolingual families. She found that the average German cry reaches its maximum pitch and intensity at around 0.45 seconds, while French cries do so later, at around 0.6 seconds.

These differences match the melodic qualities of each respective language. Many French words and phrases have a rising pitch towards the end, capped only by a falling pitch at the very end. German more often shows the opposite trend – a falling pitch towards the end of a word or phrase.

These differences in “melody contours” become apparent as soon as infants start making sounds of their own. While Mampe can’t rule out the possibility that the infants learned about the sounds of their native tongue the few days following their birth, she thinks it’s more likely that they start tuning into the own language in the womb.

In some ways, this isn’t surprising. Features like melody, rhythm and intensity (collectively known as prosody) travel well across the wall of the stomach and they reach the womb with minimum disruption. We know that infants are very sensitive to prosodic features well before they start speaking themselves, which helps them learn their own mother tongue.

But this learning process starts as early as the third trimester. We know this because newborns prefer the sound of their mother’s voice compared to those of strangers. And when their mums speak to them in the saccharine “motherese”, they can suss out the emotional content of those words through analysing their melody.

Mampe’s data show that not only can infants sense the qualities of their native tongue, they can also imitate them in their first days of life. Previously, studies have found that babies can imitate the vowel sounds of adults only after 12 weeks of life, but clearly other features like pitch can be imitated much earlier. They’re helped by the fact that crying only requires them to coordinate their breathing and vocal cord movements, while making speech sounds requires far more complex feats of muscular gymnastics that are only possible after a few months. 

Reference: Current Biology doi:10.1016/j.cub.2009.09.064

More on child development: 



A Blog by

Infants match human words to human faces and monkey calls to monkey faces (but not quacks to duck faces)


From a young age, children learn about the sounds that animals make. But even without teaching aides like Old Macdonald’s farm, it turns out that very young babies have an intuitive understanding of the noises that humans, and even monkeys, ought to make. Athena Vouloumanos from New York University found that at just five months of age, infants match human speech to human faces and monkey calls to monkey faces. Amazingly, this wasn’t a question of experience – the same infants failed to match quacks to duck faces, even though they had more experience with ducks than monkeys.

Voloumanos worked with a dozen five-month-old infants from English- and French-speaking homes. She found that they spent longer looking at human faces when they were paired with spoken words than with monkey or duck calls. They clearly expect human faces, and not animal ones, to produce speech, even when the words in question came from a language – Japanese – that they were unfamiliar with. However, the fact that it was speech was essential; human laughter failed to grab their attention in the same way, and they didn’t show any biases towards either human or monkey faces.

More surprisingly, the babies also understood the types of calls that monkeys ought to make. They spent more time staring at monkey faces that were paired with monkey calls, than those paired with human words or with duck quacks.


That’s certainly unexpected. These babies had no experience with the sight or sounds of rhesus monkeys but they ‘got’ that monkey calls most likely come from monkey faces. Similarly, they appreciated that a human face is an unlikely source of a monkey call even though they could hardly have experienced every possible sound that the human mouth can make.

Perhaps they were just lumping all non-human calls and faces into one category? That can’t be true, for they would have matched the monkey faces to either monkey or duck calls. Perhaps they matched monkeys to their calls because they ruled out a link to more familiar human or duck sounds? That’s unlikely too, for the infants failed to match ducks faces to quacks!

Instead, Vouloumanos believes that babies have an innate ability to predict the types of noises that come from certain faces, and vice versa. Anatomy shapes the sound of a call into a audio signature that’s specific to each species. A human vocal tract can’t produce the same repertoire of noises as a monkey’s and vice versa. Monkeys can produce a wider range of frequencies than humans can, but thanks to innovations in the shape of our mouth and tongue, we’re better at subtly altering the sounds we make within our narrower range.

So the very shape of the face can provide clues about the noises likely to emerge from it, and previous studies have found that infants are very sensitive to these cues. This may also explain why they failed to match duck faces with their quacks – their visages as so vastly different to the basic primate design that they might not even be registered as faces, let alone as potential clues about sound.

If that’s not enough, Vouloumanos has a second possible explanation – perhaps babies use their knowledge of human sounds to set up a sort of “similarity gradient”. Simply put, monkey faces are sort of like human faces but noticeably different, so monkey calls should be sort of like human calls but noticeably different.

Either way, it’s clear that very young babies are remarkably sensitive to the sounds of their own species, particularly those of speech. The five month mark seems to be an important turning point, not just for this ability but for many others. By five months, they can already match faces with voices on the basis of age or emotion, but only after that does their ear for voices truly develop, allowing them to tune in to specific voices, or to the distinct sounds of their native language.

Reference: PNAS doi: 10.1073/pnas.0906049106

More on child development:


A Blog by

Guerrilla reading – what former revolutionaries tell us about the neuroscience of literacy

In the 1990s, Colombia reintegrated five left-wing guerrilla groups back into mainstream society after decades of conflict. Education was a big priority – many of the guerrillas had spent their entire lives fighting and were more familiar with the grasp of a gun than a pencil. Reintegration offered them the chance to learn to read and write for the first time in their lives, but it also offered Manuel Carreiras a chance to study what happens in the human brain as we become literate.

FARC.jpgOf course, millions of people – children – learn to read every year but this new skill arrives in the context of many others. Their brains grow quickly, they learn at a tremendous pace, and there’s generally so much going on that their developing are next to useless for understanding the changes wrought by literacy. Such a quest would be like looking for a snowflake on a glacier. Far better to study what happens when fully-grown adults, whose brains have gone past those hectic days of development, learn to read.

To that end, Carreiras scanned the brains of 42 adult ex-guerrillas, 20 of whom had just completed a literacy programme in Spanish. The other 22, who had shared similar ages, backgrounds and mental abilities, had yet to start the course. The scans revealed a neural signature of literacy, changes in the brain that are exclusive to reading.

These changes affected both the white matter – the brain’s wiring system consisting of the long arms of nerve cells, and the grey matter, consisting of the nerve cells’ central bodies. Compared to their illiterate peers, the newly literate guerrillas had more grey matter in five regions towards the back of their brains, such as their angular gyri. Some are thought to help us process the things we see, others help to recognise words and others process the sounds of language.

The late-literate group also had more white matter in the splenium. This part of the brain is frequently damaged in patients with alexia, who have excellent language skills marred only by a specific inability to read.

All of these areas are connected. Using a technique called diffusion tensor imaging that measures the connections between different parts of the brain, Carreiras showed that the grey matter areas on both sides of the brain (particularly the angular gyri and dorsal occipital gyri) are linked to one another via the splenium.

Learning to read involves strengthening these connections. Carreiras demonstrated this by comparing the brain activity of 20 literate adults as they either read the names of various objects or named the objects from pictures. The study showed that reading, compared to simple object-naming, involved stronger connections between the five gray matter areas identified in the former guerrillas, particularly the dorsal occipital gyri (DOCC, involved in processing images) and the supramarginal gyri (SMG, involved in processing sounds).

 Meanwhile, the angular gyrus, which deals with the meanings of words, exerts a degree of executive control over the other areas. Learning to read also involves more cross-talk between the angular gyri on both sides of the brain, and Carreiras suggests that this crucial area helps us to discriminate between words that look similar (such as chain or chair), based on their context.

These changes are a neural signature of literacy. Carreiras’s evidence is particularly strong because he homed in on the same part of the brain using three different types of brain-scanning techniques, and because he worked with people who learned to read as adults and as children.

The lessons from this study should be a boon to researchers working on dyslexia.  Many other studies have shown that dyslexics have less grey matter in key regions at the back of their brain, and less white matter in the splenium connecting these areas. But this insights gained from the Colombians suggests that these deficits aren’t the cause of reading difficulties, they are a result of them.

Reference: Nature 10.1038/nature08461

Image: By Sgiraldoa

More on language


A Blog by

The evolution of the past tense – how verbs change over time

This article is reposted from the old WordPress incarnation of Not Exactly Rocket Science. The blog is on holiday until the start of October, when I’ll return with fresh material.

In Chaucer's time, English had many more irregular verbs than now.For decades, scientists have realised that languages evolve in strikingly similar ways to genes and living things. Their words and grammars change and mutate over time, and new versions slowly rise to dominance while other face extinction.

In this evolutionary analogy, old texts like the Canterbury Tales are the English language’s version of the fossil record. They preserve the existence of words that used to be commonplace before they lost a linguistic Darwinian conflict with other, more popular forms.

Now, Erez Lieberman, Martin Nowak and colleagues from Harvard University are looking at this record to mathematically model how our verbs evolved and how they will change in the future.

Today, the majority of English verbs take the suffix ‘-ed’ in their past tense versions. Sitting alongside these regular verbs like ‘talked’ or ‘typed’ are irregular ones that obey more antiquated rules (like ‘sang/sung’ or ‘drank/drunk’) or obey no rules at all (like ‘went’ and ‘had’).

In the Old English of Beowulf, seven different rules competed for governance of English verbs, and only about 75% followed the “-ed” rule. As the centuries ticked by, the irregular verbs became fewer and far between. With new additions to the lexicon taking on the standard regular form (‘googled’ and ’emailed’), the irregulars face massive pressure to regularise and conform.

Today, less than 3% of verbs are irregular but they wield a disproportionate power. The ten most commonly used English verbs – be, have, do, go say, can, will, see, take and get – are all irregular. Lieberman found that this is because irregular verbs are weeded out much more slowly if they are commonly used.