Grammar on the Basketball Court

The Court in Managua

Eli: Managua, 1986. A linguist named Judy Kegl has come to help with deaf education, at the invitation of the Ministry of Education. As Kegl tells it, she walked in and said, “I’m a linguist, I study sign language, what do you want from me?” The educators pointed to a bunch of kids milling around on a little basketball court. And they said, “We want to understand what they’re talking about.”

Mara: Teachers asking a visitor to translate their own students.

Eli: It’s an odd doorway into the question a listener sent us: how did humans start talking, and could we ever find out what the first words were? From that court, the story runs back into deep time and out to the edge of what linguists can recover. First, why the teachers were lost. Before the 1970s, deaf Nicaraguans had few chances to meet, so the country had no standard sign language. Around 1980, special education expanded, and schools like Melania Morales brought deaf kids together. Lessons were in spoken Spanish, mostly lip-reading, gestures discouraged. The kids signed to each other anyway. That signing became Nicaraguan Sign Language. Today's angle is the spontaneous birth of a new language in Managua.

Mara: So a whole language nobody taught them. The kids built it among themselves.

Eli: Built right under the teachers’ noses, and caught early. Kegl and her colleagues were among the first to document it. What she was looking at bears on a question far older than Nicaragua: whether a language can build its own structure, or has to be handed down from an older one.

Lessons of the Silent Kings

Mara: Those kids didn’t arrive with nothing. Many had grown up using a homesign, a private set of gestures a deaf child invents to get by with family. At school, all those private systems met. The first group, who came in during the late 1970s and early 1980s, pooled their gestures and homesigns into something ordered, with a stable vocabulary. Enough to talk about things beyond what was right in front of them.

Eli: And even one child’s homesign has vocabulary, word categories and word order.

Mara: Which is the striking part. A researcher who studies Nicaraguan homesigners says these kids, with no language around them, do it anyway. They’re just driven to, and she sees in that our natural endowment as humans for language. A team studying the new language raises the same possibility: that the capacity to create language is present in every human mind.

Eli: Kings once tested a version of that idea the hard way. Herodotus reported that an Egyptian pharaoh had two newborns raised with only goats and a mute shepherd. After two years they reportedly said ‘bekos,’ Phrygian for bread. Likely an imitation of the goats. Around 1500, a Scottish king’s isolated children were said to speak Hebrew. About a century later, a Mughal emperor had babies raised in silence. They produced no speech at all. Isolation turned up no hidden first tongue. What those children lacked was access to human speech. What the human body brought to talking, long before any school existed, is where Mara has been digging.

Bones, Brains, and Choking Hazards

Mara: For a long time, asking how talking began was nearly off limits. In 1866 the Société de Linguistique de Paris, the Linguistic Society of Paris, banned the topic outright, because speculation had run far ahead of evidence. Long afterward, as two researchers in the field put it, anyone theorizing about it risked being dismissed as a crank. What the kings and Managua both point to, a need for other people, is where later evidence picked up: changes in early humans’ bodies and social lives.

Eli: So the evidence would have to come from bones, not from arguments.

Mara: And the bones tell a story of things arriving together. A review of the fossil record finds the jaw, the chewing muscles and the vocal tract shrinking and changing alongside processed and then cooked food, the first stone tools, bigger brains, and likely a shifting social life. The authors suggest one link: once hands and stone tools did the food processing, a lighter jaw was affordable.

Eli: So the toolkit paid for the mouth that would talk.

Mara: Then there's the throat. In humans the larynx, the voice box, sits low: a descended larynx. That leaves a long chamber above the vocal folds, the vocal cords that vibrate to make sound, the pharynx, which works as a resonator for a wider, clearer range of sounds. Other primates have almost none. Peter MacNeilage, who traces language to vocal calls, points to the cost: humans can choke on food. Monkeys can't make speech sounds, but they don't get food stuck in their windpipe.

Eli: Monkeys got the safer throat. Humans got conversation, with a side of risk.

Mara: Whether talking started in the throat at all is where the readings split.

Eli: Mara, the gestural theory reads the evidence another way. Language began in the hands, as gestures and pantomime, and moved into the voice later. The idea goes back at least to the 18th century. The anthropologist Gordon W. Hewes revived it in 1973. The brain evidence on his side is the mirror system, brain circuits that fire both when performing an action and watching it. In primates, it represents hand gestures and some mouth movements, but not calls.

Mara: On Hewes's reading, the hands came first, and the voice inherited the job.

Eli: Chris Petkov reads it the other way: speech grew out of primate calls, and he has brain evidence of his own. In 2020, his team found a segment of the auditory language pathway, wiring that connects the brain's hearing areas to frontal regions that handle speech. They found counterparts of it in apes and monkeys. Petkov said, "I admit we were astounded to see evidence for evolutionary homologs of the human language pathway in both apes and monkeys, linked by way of an auditory origin."

Mara: Homologs: the same wiring, inherited from a shared ancestor. So both camps reach deep into the primate past, the hands through a system for imitation, the voice through a pathway apes and monkeys share. The fossils hold the hardware. How that hardware ever came to carry grammar is something bones can’t show.

Syntax Takes the Yard

Eli: Kegl noticed a split. The older students used an elaborate gesture system. The younger ones, arriving year after year, did something else. They broke whole gestures into separate pieces with rules for recombining them: discrete combinatorial grammar, the way words slot into a sentence. With the second group, in the late 1980s, came spatial grammar, placing people and things in the space around the body to show left and right, and who did what to whom. Even a furrowed brow became grammar.

Mara: So the youngest outpaced the founders who’d pooled the homesigns in the first place.

Eli: And Kegl saw it plainly. “When you see a language, it’s rule-governed. It’s got internal consistency,” she said. Watching these kids, “it was like somebody just wiped away the fog, and you could see the grammar right there in front of you.” Numbers tightened too. Before 1990, six meant holding up six fingers, and twenty meant signing ten twice. By the early 1990s, those had become small, fast one-handed signs.

Mara: Language comes from exposure plus inborn capacity; thin input makes the inborn side show.

Eli: That's the turn. Kegl was asked to translate a playground. What she found was grammar forming among children, cohort by cohort, not inherited from some ancient mother tongue. What no one has pinned down is when, in which ancestor, or what first turned early human calls into structured speech.

Mara: Meanwhile, deaf kids outside the capital were still growing up without schooling.

Eli: In 1994, Kegl’s husband and a deaf anthropologist stomped through hamlets around Bluefields and found about seventeen deaf people. Their parents had formed a group for kids with disabilities, but the children had no real schooling. When deaf signers came from Managua, the kids followed them out into the backyard like the Pied Piper. An interpreter told the parents that language would let them communicate with their children instead of relying on physical discipline. That’s language dissemination here: deaf adults passing it on, person to person.

Mara: One later teacher there, born deaf, barely communicated until starting school at four.

Eli: She started at the Bluefields school, learning partly by matching pictures to signs. After high school she joined traveling teachers on Ometepe Island, learning deaf islanders' homemade signs and teaching them the new language. Kegl came to translate one schoolyard; its language spread well beyond it. Whether anyone could work backward the same way, to the first spoken words, is what Mara has been chasing.

The Ten-Thousand-Year Ceiling

Mara: Working backward is what historical linguists do. Their main tool is the comparative method: line up related languages, find the regular sound matches between them, and rebuild the ancestor they share. It works, up to a point. Words keep getting swapped out over time, which linguists call lexical replacement, and each swap erases a clue. Johanna Nichols puts the practical limit at about eight thousand years, or a bit more. She treats that as a ceiling. How hard a ceiling it is, linguists still debate.

Eli: The other side thinks the ceiling can be broken. Merritt Ruhlen argues that every spoken language descends from one universal tongue, often called Proto-World, a single ancestor for them all. He proposed a set of global roots, each a sound shape with a broad meaning. Another team, working on giant language groupings, says some shared words aren't chance look-alikes and can be rebuilt deep in time. They provisionally joined four huge groups into one super-family called Borean. Most linguists doubt that all languages share a common origin.

Mara: Louis-Jean Boë and colleagues put Ruhlen's roots to a test, with a simulation of random languages. The key was how loose each root was. Ruhlen let a root cover about two dozen meanings, on average. In their model, a root with just a handful of meanings already turned up matches every time, by pure chance. Their conclusion: chance can't be ruled out. So, the listener's first question. The question Paris banned now has bones to work with: tools, cooked food, a lighter jaw and a rebuilt throat. Modern speech anatomy and spoken forms took shape around a hundred thousand to fifty thousand years ago.

Eli: Far older than anything the comparative method can reach.

Mara: And Managua adds a twist. Signed languages invented by deaf communities share speech’s essentials: reference, open-ended combination, grammar, even prosody, the rhythm and emphasis of a sentence. Language doesn’t depend on the voice; that’s modality independence. So whether hands or voice came first, the machinery of grammar isn’t tied to either. And one origin for our species doesn’t have to mean one original language. Those questions stay separate.

Eli: Which is what the kids on that court showed, without a single sound.

Mara: So the plain answer. Humans began talking through a slow combination of vocal, manual and social changes. The first words are out of reach for good, because replacement erases vocabulary beyond roughly ten thousand years. Kegl was asked what some kids on a basketball court were saying. There, researchers could watch grammar arrive. For the first spoken words, there’s no court to walk onto.

Eli: Three things could still move this. Ancient DNA or work on gene regulation could pin down when modern vocal control evolved. Computational comparisons could show words surviving past that ceiling more often than chance allows. And imaging primate brains could settle whether speech grew from gesture or from calls.

Mara: Dig into Nicaraguan Sign Language.

Eli: And stay with us on science.

Mara: That's it for today. I'm Mara.

Eli: And I'm Eli. Thanks for listening, and until next time on Angle.