No Return Ticket: When Languages Reach the Point of No Departure

As we move through this decade of focus, a greater understanding of the mechanisms of language decline is necessary. The discussion must move beyond the classic narratives of attrition to examine the finer relationship between state power, community agency, and the very identity of a people.

The Anatomy of Language Loss: A Process, Not an Event
The notion that languages fade away through a form of "natural selection", because they are no longer needed for communication, is an oversimplification that obscures what is really happening. A more precise framework distinguishes between external actions and internal responses. The term "linguicide" describes the deliberate actions, whether overt or covert, that lead to the death of a language. This process is rooted in "linguicism", a system of ideologies and practices that legitimizes an unequal distribution of power and resources between groups based on language, akin to racism.

However, the final decision to cease transmitting a language to the next generation often rests with the speakers themselves — a phenomenon termed linguistic suicide. These two concepts are not mutually exclusive. A more integrated model suggests that linguicide creates the conditions under which linguistic suicide becomes a rational, albeit tragic, choice. Prolonged exposure to linguicidal ideologies can trigger an attitude shift within a community, where speakers internalize the negative perceptions of their own language and come to see its abandonment as a pragmatic step toward social and economic opportunity.

The consequences of this process are severe. Among the Pijao of Colombia, whose language was eradicated through overt government policy, older generations recall the violence and persecution that accompanied its use. The resulting shame felt by elders has been partly responsible for the non-transmission of the few remaining words, leading to a progressive weakening of Pijao identity across generations to the point where younger members struggle to define what makes their group distinct.

The Role of the State: From Suppression to Ambivalent Support
Government policy is a primary instrument of linguicide. These policies can be categorized by their approach, ranging from active hostility to passive neglect. Active Linguicide consists of overt measures to eradicate a language. The French Republic, for example, has a long history of this. Following the French Revolution, the centralizing state pursued a monoglot ideology to forge a unified nation. Regional languages, or patois, were demonized as barbaric jargons and symbols of counter-revolutionary sentiment. And the education system became a key tool of suppression. In schools across France, children were punished for speaking their mother tongues, with practices like the symbole — an object of humiliation passed to any student caught speaking a regional language — designed to instill a sense of shame and inferiority. Rules explicitly forbade "speaking Breton and spitting on the ground", a practice that directly linked the language to unhygienic behavior. The symbole was transposed in the French-occupied country known as French Polynesia, with students shamefully passed on a black cowrie shell big enough to be visible by all.

Passive Linguicide, while less overtly violent, can be equally destructive. This includes benign neglect, where a state allows a language to decline by refusing to provide support. A more subtle form is folklorization, where a language is given symbolic value as cultural heritage but denied a functional role in modern society. France's 1951 Loi Deixonne, which permitted the teaching of local languages and dialects alongside folklore, literature and local folk arts, in practice positioned these languages as artifacts of a bygone era rather than living tools for contemporary life. Similarly, the 2008 constitutional amendment stating that regional languages belong to the patrimony of France has been legally interpreted as an expression of attachment that creates no enforceable rights for their use in official capacities.

In contrast, some nations have adopted proactive frameworks. The Welsh Language Measure of 2011 conferred de jure official status upon Welsh in Wales, obligating public bodies to provide services in the language. The United States' Native American Languages Act of 1990 formally recognized the right of tribes to use their languages as a medium of instruction, a direct reversal of decades of assimilationist boarding school policies. These examples show that state policy can also be a powerful force for revitalization.

A View from the Pacific Islands
The Pacific presents its own distinct set of circumstances. Within the political entity known as the Realm of New Zealand, which includes the Cook Islands, Niuē, and Tokelau, a telling disparity exists between the languages of its associated states and those of independent Pacific nations. While Samoan and Tongan maintain a degree of vitality in Aotearoa, the languages of these associated states — Rarotongan, Niuean, and Tokelauan — are however severely endangered.

Data from Aotearoa's censuses illustrate this decline. In 2023, speakers of a heritage language from these associated territories comprised a mere 3% of the Pacific population in New Zealand, a fall from 5% in 2013. In contrast, Samoan speakers stood at 24% in 2023. This is not an accident but a direct result of New Zealand's administration. The state has enacted linguicide on the languages of its associated territories primarily through the education system, which, even in the island nations themselves, largely follows the New Zealand curriculum and prescribes English or Māori as the language of instruction. This has resulted in a situation where, for instance, over 90% of children in Rarotonga and Niuē arrive at school dominant in English. The result is that communities from New Zealand's associated states are far more advanced in their shift to English than Pacific communities from independent home countries like the Kiribati.

Technology, Data, and the Path Forward
Revitalization is a multifaceted endeavor, categorized by the language's current status: Reclamation for languages with no living speakers like Hebrew (here, Hebrew as opposed to Ivrit), revitalization for severely endangered languages like Káruk in California or Navaho in the Navajo Nation, and reinvigoration for languages with a substantial speaker base but restricted use like Belarusian.

In the digital age, technology is often heralded as a key tool. E-learning platforms like Duolingo have substantially increased the number of learners for languages like Welsh, Irish, and Scottish Gaelic and now provide unprecedented access. However, for many of the world's languages, especially those of the Pacific, the large-scale public data needed to build such tools is nonexistent. This is a domain where specialized, culturally informed work becomes indispensable.

At our firm, we recognized this data deficit early on. Since 2008, we have undertaken the meticulous work of curating our own datasets for the Pacific Island languages we serve. This has allowed us to develop language models that are not publicly available — At least for now. These models assist our human teams of cultural and linguistic experts in a range of tasks, from translation and editing to proofreading and subtitling. And this work is grounded in the practical domains where these languages are strongly used today: user interfaces for digital tools, healthcare communications, immigration documents, seasonal work contracts for the meat processing industry, and military contexts — while also supporting projects related to cultural heritage to a lesser extent. This approach counters the trend of folklorization and museumization, and allows our languages to retain their utility in modern, high-stakes environments.

The UN Decade of Indigenous Languages is a call for concerted action. The evidence from around the world shows that language loss is not inevitable. Far from it. The path is not uniform, however. in fact, success depends on a combination of factors: Genuine state support (not just folklorisation or museumisation), community-led initiatives, and the intelligent application of technology. The central question facing us is not so much if we can save these languages, but rather what kind of future we are saving them for: One where they are revered as relics in a museum, or one where they thrive as living, evolving expressions of human identity and knowledge in a modern world.