How much does our knowledge of obscure languages depend on missionary work which preserved and exposed them?

Quite a bit.

I trained around fieldwork linguists. Which was a colossal mistake for someone working on a European language. But useful if you want to be exposed to typology. I hear the IPA horror stories of my peers here, and blanche. Can linguists differentiate between all the sounds of the IPA?

Now. Fieldwork linguists tend to like the cultures of the people they work among. And they tend not to like people they perceive as working to eradicate those cultures. They are academics, so they’re already on preponderance predisposed against religion. (The most visionary linguist we had was a pastor in training, and eventually moved across to theology—and anti-Islamic rhetoric. Our homegrown computational linguistics giant was a fieldworker, and a missionary. Both were regarded by their colleagues with amused detachment.)

I learned more about typology working as a research assistant than doing a PhD. Just as well, because my PhD wasn’t in typology. I was research assistant for a prof doing a phonological survey of Papua New Guinea.

PNG has, what, 1/8 of the world’s languages? 1/6?

How many academic linguists have signed up for a lifestyle of malaria and dysentery, so they can record obscure languages there?

More than there used to be. We’ve run out of new languages among our Indigenous Australians, so PhDs wanting to write a grammar of an underdocumented language are sent to PNG now. But still, of the 800-odd languages there, I’d be surprised if even more more than 50 are documented to an acceptable scholarly level.

You know who’s prepared to sign up for a lifestyle of malaria and dysentery, to record languages in PNG for their own motives?

Of course you do.

Academic Linguists owe a lot to SIL International. Some of them wish they didn’t. While the SIL missionaries are sometimes linguistic incompetents, sometimes they are brilliant linguists; it’s mixed.

I know one thing, though. Vanuatu has the highest density of languages on earth: 110 languages, population of 300k. Vanuatu, for reasons that a lot of linguists would be sympathetic to, has barred SIL from working there.

And Vanuatu, not PNG, is the most linguistically underdocumented country there is.