A language can disappear even while its community grows. This paper asks what the science of disappearing species can teach us about protecting the words, stories, and knowledge carried by human languages. Language diversity is under threat: between a third and a half of all languages are considered endangered, with predicted loss equivalent to one language per month for the rest of the century.

The causes of language loss differ from the causes of species loss, but studying both kinds of diversity creates similar challenges: testing explanations and identifying causes. The review asks how tools developed to study species diversity can be applied to language endangerment and loss.

Many methods from biology can be adapted, as long as the similarities and differences are both recognised. Broad studies cannot capture the local processes behind language loss, but they can reveal general patterns and possible contributors to the decline of linguistic diversity.

For species, one shared system provides a global standard for judging endangerment. Language has no equivalent internationally agreed standard, and measures based on population size and range decline do not fit languages well.

Instead, language scales focus on whether a language passes from one generation to the next and where it is used. Several databases use different scales and cover different sets of languages. The scales broadly agree, but they can rank individual languages differently.

That means a language may receive different warnings depending on which database is consulted. So they tried something clever: before asking whether a condition is linked to language danger, they account for the fact that related languages often resemble one another.

It is like comparing family recipes: several relatives may use the same ingredients because they learned from one another, not because each independently discovered the same recipe. Language relationships can help separate shared history from separate evidence.

Finding factors linked with endangerment can help predict broad future patterns of language loss. Conditions such as climate change or changes in land use can be used to project future pressures on languages. This raises the possibility of identifying languages that are not currently in danger but already face conditions that could place them at risk.

But here's the catch: some forecasting tools from conservation biology have limited use for languages. A method based on birth and death rates works best when population growth can be carried into the future reliably. Language vitality depends on transmission, not human reproduction, so the number of speakers may not follow the size of the population.

In one Australian case, the speaker number declined for over a century while the population grew, and the language became critically endangered. The headline finding is not that biology and language are identical. Their patterns of diversity and loss can look strikingly similar, but biological tools should be used only when they capture processes that matter for language.

Broad analysis cannot tell the whole story of any particular language, yet it can reveal patterns of loss and point toward important interventions. For people trying to keep a language alive, the practical lesson is direct: broad-scale evidence can guide attention, but it cannot replace local understanding of its vitality and future.

One important intervention is education that supports language diversity rather than eroding it, helping communities pass their languages on instead of pushing them aside. Biology offers useful ways to find broad patterns in language loss, but no single measure can replace local knowledge.

The practical goal is education and policy that support, rather than erode, language diversity.