Linguist and conservative commentator John McWhorter estimates the 6,000 languages spoken today will dwindle to only 600 next century. He argues that this is part of a process that will confer economic and health benefits to the affected speakers.
His main point is that the vast, vast majority of threatened languages are those spoken by isolated indigenous groups, and that these languages are, in fact, a driving force of their isolation. The language barrier prevents the absorption of such groups into the larger society, and this often leaves those affected in significantly worse economic conditions than their neighbors that speak the majority language.
McWhorter outlines how the pursuit of a better life can often mean leaving one's ancestral language behind:
As people speaking indigenous languages migrate to cities, inevitably they learn globally dominant languages like English and use them in their interactions with one another. The immigrants' children may use their parents' indigenous languages at home. But they never know those languages as part of their public life, and will therefore be more comfortable with the official language of the world they grow up in. For the most part, they will speak this language to their own children. These children will not know the indigenous languages of their grandparents, and thus pretty soon they will not be spoken. This is language death.
The controversial part is where he questions the importance of keeping endangered languages alive. To be sure, he feels languages should be recorded and preserved, something for which modern technology thankfully allows, but he questions the wisdom of investing huge amounts of financial and human resources in ensuring groups continue to speak the language of their ancestors. Many such languages are extraordinarily difficult for non-native speakers to learn, which can hugely complicate the task of professional linguists who try to teach these languages. When the main motivation to keep a language alive is a relatively abstract, aesthetic one, it may prove impossible to turn the tide on language death. McWhorter compares the task to stopping ice from melting.
He does acknowledge, however, that there are good historical reasons for people to feel wary of this process, including the growing universality of English:
Obviously, the discomfort with English "taking over" is due to associations with imperialism, first on the part of the English and then, of course, the American behemoth. We cannot erase from our minds the unsavory aspects of history. Nor should we erase from our minds the fact that countless languages-such as most of the indigenous languages of North America and Australia-have become extinct not because of something as abstract and gradual as globalization, but because of violence, annexation, and cultural extermination. But we cannot change that history, nor is it currently conceivable how we could arrange for some other language to replace the growing universality of English.
He also takes some time to consider the argument that languages should remain alive because they encode unique cultural worldviews. He suggests that, although languages are obviously key features of what make cultures distinct, it is a mistake to overstate how much they encode thought patterns unique to its speakers.
To illustrate this point, he considers an appraisal of the recently extinct Alaskan language Eyak:
One school of thought proposes that there is more than mere chance in how a language's words emerge, and that if we look closely we see culture peeping through. For example, in its obituary for Eyak, the Economist proposed that the fact that kultahl meant both leaf and feather signified a cultural appreciation of the unique spiritual relationship of trees and birds. But in English we use hover to refer both to the act of waiting, suspended, in the air and the act of staying close to a mate at a cocktail party to ward off potential rivals. Notice how much less interesting that is to us than the bit about the Eyak and leaves and feathers.
As someone with a smattering of linguistics background, I'm not necessarily convinced by everything McWhorter puts forth, but his paper is well worth reading for a somewhat heterodox perspective on language death and how it will fit with our other cultural priorities in the next hundred years.