ExplWorld
Dispatch · Papua New Guinea

Taught in Four Hundred Languages

Papua New Guinea has around 840 living languages, more than a tenth of every language alive on earth, in a population of about ten million. In 1995 the state decided to teach its six-year-olds in them, by the early 2000s more than four hundred languages were in use in its classrooms, and in 2013 it stopped.

ExplWorld Editorial
7 August 2026 · 6 min read · Vol. 1 · Summer 2026

Ethnologue counts 840 living languages in Papua New Guinea. That is more than any other country and something over a tenth of the world total, spoken by a population usually given as around ten million, on an island where a day's walk between two valleys can put you among people whose grammar shares nothing with your own. The reason is topography and the reason is time: the Highlands have been farmed for thousands of years by communities that had every reason to stay where their gardens were, and mountains do to language what oceans do.

What a country does about that in a classroom is a real question with no comfortable answer, and Papua New Guinea has answered it twice in opposite directions inside thirty years. From 1995 it taught the first three years of school in whatever language the village spoke. From 2013 it taught them in English. Both decisions were taken by governments that said they were acting for the same children.

The schools that started it

The idea did not come from Port Moresby. In 1980 North Solomons Province — Bougainville — set up what it called Viles Tok Ples Skul, village vernacular schools, giving children two years of education in the local language before they entered Grade 1 in English. SIL had trained provincial staff in producing literacy materials there the year before. Between 1981 and 1985 two more provinces and four more language communities did the same, East New Britain among them, and the results were mixed enough to argue about and good enough to copy. This was a provincial improvisation that the national government eventually adopted, not a plan handed down.

What the reform required

An education sectoral review in 1991 produced a reform proposal, parliament approved it in 1993, pilot elementary classes ran in Milne Bay and New Ireland in 1994, and the Education Reform Agenda was approved in 1995. The school structure went from six years primary, four secondary and two of upper secondary to three, six and four: three years of elementary school — Prep, Elementary 1, Elementary 2 — taught in the language of the community, followed by a bridge into English through primary. The mechanism that made it conceivable was that the community chose its own elementary teacher, who was trained locally and taught in a language they already spoke. Nobody in a ministry had to find four hundred qualified linguists. By the early 2000s more than four hundred languages were in use in the formal school system, which is a scale of mother-tongue instruction no state has attempted before or since.

A state with 840 languages cannot order textbooks. It can only ask every village to write its own, and then find out how many are able to.

What broke

Delivery, not principle. Most of those languages had no agreed orthography, so there was nothing to print and nothing to print it on; teaching materials had to be made by the elementary teacher, by hand, in a building that frequently had no floor. Those teachers had themselves been schooled entirely in English and many were uneasy with methods they had never been taught. Professional support was thin and arrived unevenly — almost all the early programmes leaned on SIL, whose personnel were seconded for a decade to help provinces with materials and training, and which could not be in every province at once. In towns, where no single language dominated, "the language of the community" meant Tok Pisin in practice, which was not what the reform's architects had in mind. And parents, who understood perfectly well that English was the language of the civil service, the mine payroll and the university, watched their children spend three years not acquiring it.

The reversal

In August 2012 the prime minister, Peter O'Neill, announced that outcomes-based education would be scrapped. On 22 January 2013 the National Executive Council appointed a task force to examine it; the task force recommended replacing the outcomes-based curriculum with a standards-based one, and from 2013 elementary instruction reverted to English, with English also taught as a subject. The government's stated reason was that vernacular instruction had produced the country's poor standard of English — and that reason is the part worth stopping on, because the comparative evidence points the other way. Work in East New Britain found that children taught first in Tinatatuna, the Tolai language, did better at literacy and numeracy than children taught only in English, which is the ordinary finding of mother-tongue-first research more or less everywhere it has been done. What was failing in Papua New Guinea was not the theory of the reform but its supply chain. The structure has since changed again, from three-six-four to one-six-six: one preparatory year, six of primary, six of secondary, introduced from 2021 with full rollout targeted for 2024, and in early 2025 the government signed a compact with seventeen partners to finish a job that is still not finished. English remains the medium from the first day of school.

What the counting shows

The best measurement anyone has taken was published in PNAS in 2021. Between April 2015 and November 2018, researchers tested 6,190 upper-secondary students speaking 392 languages at 30 of the country's 123 secondary schools. Fifty-eight per cent of the students were fluent in an indigenous language, against 91 per cent of their parents, and the drivers of that decline projected fluency down to about 26 per cent in the following generation. The same students knew fewer bird species and fewer plant uses than their parents did, and the two losses tracked one another closely, because the names and the knowledge travel together. The finding that ought to unsettle both sides of the schooling argument is that the students' English and mathematics scores had no bearing on their indigenous-language skills at all. What was changing them was not the classroom. It was everything else.

On a bus out of Goroka, in the Mount Hagen market, at a Sepik landing where a canoe is being loaded, the language doing the work is Tok Pisin. It is nobody's ancestral language and very nearly everybody's second one, it is the reason a trip like this is possible at all, and about two hundred words of it will change every transaction you have. The 840 languages underneath it were never held up by policy. They were held up by distance, by valleys a day apart, and by grandmothers who spoke no other — and roads and phones are dismantling the first two of those faster than any ministry in Waigani could either help or hinder. Which does not make the argument about what happens in the first three years of school pointless. It makes it the only part of this that a government still controls.

Featured countries