In a classroom on the main floor of the Mashpee Wampanoag tribal resources center, a dozen or so toddlers are speaking a language that until a few decades ago had no living speakers.

This is the language of the Wampanoag Nation, the first language of Massachusetts. The fact that these children are learning it is due to a movement that began with one woman, Jessie Little Doe Baird, whose office is down the hall. Baird's success at creating a layperson's handbook of grammar and compiling a working dictionary was unprecedented. "It's about fixing what happened," Baird said of her life's work. "It's about making whole what was broken."

In articles about Baird, and there have been many, her feat is usually portrayed as a completed triumph. The stories usually talk about how Baird won a MacArthur "genius" grant in 2010. They describe the Wampanoag language — called Wôpanâak or Wôpanâôt8âôk in its own modified alphabet — as "resurrected." In the global movement to preserve and restore endangered languages, the tribe's Wôpanâak Language Reclamation Project is a "poster child," said linguist Lenore Grenoble.

Except languages don't live in dictionaries; a language is truly alive only when spoken in a community. To Baird and her co-speakers, Wampanoag is a living being, to be nurtured and protected. For years, the tribe did not allow non-Wampanoag to learn it, and still does not teach it to people who aren't closely tied to the Wampanoag community. "You can't say I reclaimed anything," says Baird. "It's not humanly possible for one person to reclaim something that's social in nature."

Those whose ancestors have always spoken English, the dominant tongue of global commerce, might not grasp how important it is for a language to have even a few fluent speakers, rather than none. Linguists estimate that more than 40 percent of the world's languages are vulnerable to extinction within a generation. Around the world, indigenous people are fighting to preserve their dwindling languages. Reviving a language that had no native speakers left, as in the case of Wampanoag, is still harder. The barriers Baird and her collaborators have encountered — beyond the logistical challenges, there are moral and ethical ones — typify the difficulties facing all the world's threatened languages.

Three decades on, Wampanoag has come further than Baird ever thought possible. But it's still a long way from being a fully fledged, full-time community language. There is more at stake here in this preschool classroom than the fate of Wampanoag, or even the self-determination and sovereignty of the tribe to which most of these children belong. The extent to which the Wôpanâak Language Reclamation Project succeeds — and what happens to the language during that process — shows us the extent to which the extinction of languages can possibly be halted, or reversed; the extent to which a culture ransacked by genocide can recover part of its heritage.

Jessie Little Doe Baird. (Getty Images for the Macarthur Foundation)

THE PROJECT to reclaim Wampanoag was born in 1992, when Baird began dreaming in a language she didn't know.

The dreams went like this: She was walking near the sea, in a village that had been razed to cinders. Safe in a room underground, survivors who looked like her family were talking and singing. She listened to their voices but she didn't understand their words.

Baird is a practical, matter-of-fact person — the kind who knows how apply for a grant, diaper a baby, and hunt quahogs by feeling along the sand with her feet. So she knew a sacred dream when she had one.

She consulted with tribal elders, and they agreed that the dreams were a sign: It was time to see if the community wanted to "bring the language home."

Baird never speaks of Wampanoag as having "died." Instead, she prefers to say it was asleep or lost. But it hadn't wandered away. What happened to Wampanoag was an act of violence, a cruel chapter of Massachusetts history that is rarely discussed.

In the 18th century, hundreds of Wampanoag families were coerced into predatory loans, then forced to give up their children as repayment — a form of debt slavery called "indenturement." Toddlers as young as age 2 could be taken to live and work in white households to pay their parents' debt. They would return home years later, unable to communicate with their elders.

With each generation, the number of children raised away from their parents increased, and the number of adult speakers diminished. By 1907, a linguistics professor named J. Dyneley Prince was pleased to report that an informant had extracted from "unintelligent and aged persons" in Mashpee "at least twenty-nine isolated words of the old speech."

By the time Baird dreamed of her ancestors, the only words in their language she knew were a few slang terms and the place names on street signs.

This loss was more than incidental; it became a weapon used by the government against the tribe. In the late '70s, when Baird was about 12, the tribe took the town of Mashpee to court over 11,000 acres of land. The town's response was as simple as it was devastating: The land could not belong to the Mashpee Wampanoag Tribe, its lawyers argued, because there was no such thing as the Mashpee Wampanoag Tribe. After all, the town's lawyers argued, they didn't even have a language of their own.

BAIRD MIGHT NOT have seemed like the most likely candidate to revive a moribund language. She had no college degree and no particular experience with the field of linguistics.

But she wasn't one to ignore a clear spiritual mandate. She began doing research on her own, and eventually found her way to a special fellowship program at MIT. Her initial work there was so promising that she was admitted as a doctoral candidate at the school's prestigious linguistics department.

Baird's arrival at MIT coincided with a rising swell of language revitalization movements around the world. Sparked by the civil rights movements of the 1960s, efforts to strengthen indigenous languages had spread among indigenous peoples from continent to continent.

In 1973, speakers of Hualapai, in Arizona, began working to create what would become the local school district's official bilingual program. The same decade, classes in Native Hawaiian began to be offered at Hawaii universities. In 1983, the first school taught in the language of the Ainu, Japan's indigenous people, opened in Hokkaido. In the late 1980s, the Maori of New Zealand began to establish "language nests" where young children could interact with elders using the Maori language.

There were many more, from Ireland to Australia, Cornwall to California. Many of the languages had been all but extinguished through genocidal violence similar to what the Wampanoag experienced. For so many indigenous people, restoring their languages was almost an act of reverse-entropy — a way of undoing loss and affirming their identity.

By the 1990s, the field of linguistics was finally catching up. The year before Baird came to MIT, the linguist Ken Hale — who would become her friend and mentor — published a paper titled "On Endangered Languages and the Safeguarding of Diversity" in the journal of the Linguistics Society of America. In it, he argued that linguists should do more than study dying and extinct languages; they had a duty to preserve and revive them. He had reason to know: A specialist in endangered languages and a prodigious polyglot, he had found himself the sole speaker of more than one language when his last informants died of old age.

"When you lose a language," he famously told a reporter, "you lose a culture, intellectual wealth, a work of art. It's like dropping a bomb on a museum, the Louvre."

In Hale, Baird found an ideal partner. And Baird's ancestors had left them a gift: Many 17th-century Wampanoag had adopted literacy, inventing an orthographic system based on the Roman alphabet but adapted to the sounds of their language. They had left behind a huge body of documents — not only a century's worth of handwritten letters and legal documents, all in Wampanoag, but also an entire Bible, translated in the 1600s by Wampanoag converts. Poring over this treasury, Baird was able to piece together a vocabulary; she knew how to pronounce the words through comparison with related living languages.

She taught herself to speak by talking back to her own tape-recorded voice.

AS WAMPANOAG began to awaken, a touchy question arose: Who should learn the language? For years, Baird's classes were open only to members of the Wampanoag community. Any publication that uses Wampanoag words has to be vetted by the community of speakers, even if it's written by a Wampanoag. The reclamation project's website tells curious visitors that "pumpkin" and "skunk" come from Wampanoag words that translate as "grows forth round" and "ejects bodily fluids" — but reveals little about how the language looks or sounds.

The group was protective in part because the members wanted to reserve their scant resources; they had to train enough fluent speakers to become teachers in their own right. But they also wanted to protect the language from appropriation. The website also has a list of things the organization won't provide translation services for, including: "Names for your non-native child; names for your non-tribally owned property/home (especially if there is evidence of Wampanoag ancestral remains); names for your pet."

But this was meant to be a living language, not something kept in a display case. And so in the summer of 2014, as they worked on a proposal for a new charter school to be taught entirely in Wampanoag, the team members on the Wôpanâak Language Reclamation Project found that they had to construct new words for items new to the environment, such as "telephone," "computer," or "texting." (Ultimately, Baird's group joined forces with an existing Montessori school.)

Nitana Hicks Greendeer is one of Baird's most accomplished proteges, who followed in her footsteps to MIT. Her thesis project was a list of all the roots of Wampanoag words and their meanings: One might mean "red," another "thing that grows in a vine along the ground," and the two together can mean "watermelon." Whenever the group needed a new Wampanoag term for the curriculum it was working on, the members would pore through Greendeer's thesis, fitting roots together to build a new word.

The choice not to borrow any English words caused some controversy within the group. "Even in Mashpee, there's a major difference between people who think, 'If it's going to be a language, it's going to be the language,' and people who think 'No, you should borrow things,'?" Greendeer said.

Other indigenous languages have faced the same dilemma. In Cheyenne, for example, Cheyenne-speaking elders were consulted to create new words. But the Wampanoag had no fluent elders to consult. So as they built new words, they tried to stay as close as possible to the outlook of their ancestors.

For example: the word for "telephone."

Wampanoag had ceased to be spoken before telephones were invented. But if it hadn't, what aspects of a telephone would be most salient to the minds of Wampanoag speakers? What particles of meaning would construct the new word? And for that matter, what kind of telephone would these hypothetical Wampanoag-speaking ancestors have first encountered? Not a smartphone, they decided, but something with prongs and long straight wires, something only a little evolved from a telegraph.

To create the Wampanoag word for telephone — it translates as "talking line" — Baird and her group imagine a hypothetical timeline: an alternative universe in which the Wampanoag people were never forced to indenture their children, never stopped speaking their language, were never told by a 20th-century jury that they did not, in fact, exist.

ANY LANGUAGE that has undergone efforts at "revival" — even if it still has living speakers at the time — also undergoes significant change.

All languages, of course, are changing all the time. But the question of change versus "historicity," authenticity, or purism is a fraught issue in some language revitalization movements. And though it's not well studied, many revitalized languages undergo a period of rapid evolution as younger speakers adapt the old language to their own needs.

"There isn't a lot of research of that end product in terms of the language," said Grenoble, who studies endangered languages at the University of Chicago. But, she said, "You're not getting the language you would get if you had uninterrupted transmission. The language would change [anyway] — we aren't speaking the English of our ancestors. But it would change in one way, and now with interrupted transmission it will change in another way."

The most well-known language to be revived is Hebrew — which, like many indigenous revitalization efforts, was also part of a bid for identity and self-determination by a people who had survived a genocide. But the Hebrew spoken in Israel today differs significantly enough from its liturgical sources that some linguists have argued that it is a constructed language, like Esperanto or Elvish.

Other revived languages also show rapid transformations: simplification of pronouns in Gaelic, pronunciation shifts in Maori. In some cases, older speakers have worried that what has been revived is an altered or adulterated version of the "real" language. But Grenoble says it's best to accept these "impurities" if revitalization efforts are to be successful.

"You get what you get," Grenoble said. "If you insisted on having some pure Wampanoag. not only would it be impossible, it would be the kiss of death to the program."

The children in Baird's preschool are the group's best hope to make their language once again a communal tongue, transmitted naturally from one generation to the next. Children soak up languages like sponges, their brains optimized to do the work that becomes arduous at an older age. But they will also change the language in ways that can't be predicted.

Baird said she welcomes that.

"It won't be the record that was preserved, because it's going to be used," she said. "It's a living thing — it might develop six heads. If six-headed [Wampanoag] is what we end up with, well, OK."

It's in the next generation's hands, she said.

"It's always children that perform that function of language evolution ," Baird said. "We don't have to worry about it. The children will do it."

S.I. Rosenbaum can be reached at si@arrr.net. Follow her on Twitter @sirosenbaum.