Chéon in 1907, Maspéro in 1912 and Cuisinier in 1948 all filed Nguồn under Mường. Mạc Đường in 1964, Nguyễn Dương Bình in 1975 and Phạm Đức Dương the same year filed it under Vietnamese. Later comparative work by Nguyễn Văn Tài and Nguyễn Phú Phong swung back toward Mường, and Chamberlain and Sidwell eventually stopped choosing sides, counting Nguồn as a third Viet-Mường language in its own right. Some linguists have said the quiet part aloud: that the scholars placing Nguồn inside Vietnamese were guided less by sound changes than by ethnography and politics. A language classified as a dialect of the national language belongs to a people who are already the majority. A language classified as something else belongs to a people who are not.

Nguyễn Văn Tài described a demonstration that needs no laboratory. Put a Nguồn speaker and a Mường speaker together, let each use their own language, and they will manage a conversation. Put a Nguồn speaker with a Vietnamese speaker who knows no Mường, and the Vietnamese speaker will not follow. That asymmetry is the strongest single piece of evidence in the whole dispute, and it points away from Hanoi. Yet Nguồn keeps producing counter-examples like the loss of chẳng, which lines it up with Vietnamese instead. Edmondson, Gregerson and Nguyễn Văn Lợi called the language of great interest to anyone studying Vietic history for exactly this reason: it refuses to sit still. There are even two varieties pulling in different directions - Cổ Liêm, named for a village, and Yên Thọ, named for a farming cooperative in Tân Hóa, with Yên Thọ leaning further toward Vietnamese in its sound system.

In 1905 the missionary-ethnographer Léopold Cadière recorded the Nguồn living in eleven villages along the valleys of the Nguồn Năn river, in two clusters of five. The northern group - Qui Đạt, An Đức, Ba Nương, Thanh Long, Tân Kiều - lay in Cơ Sa canton alongside Việt villages, and Tân Kiều later split to make a sixth, Tân Hợp. The southern group ran Kổ Liêm, Bốk Thọ, Kim Bãng, Tân Lí, An Lạk. Most of that ground is now in Minh Hóa district of Quảng Bình province, a district of 1,401 square kilometres and about fifty thousand people whose administrative seat is Quy Đạt - one of Cadière's eleven. Others live around Đồng Lê in Tuyên Hóa district, roughly 50 km off National Highway 1, sharing the hills with the Sách, a Chứt group whose own language is Vietic too.

Where the Nguồn came from is contested along the same fault line. One account holds they are a Việt population out of Hà Tĩnh and Nghệ An who moved south into these valleys by the seventeenth century, an argument built on family records. The other holds that Mường groups had already settled the area, that later Việt arrivals assimilated into Mường speech, and that contact with Chứt languages did the rest. What is not contested is the paperwork. On the 1960 census, most Nguồn declared themselves Việt. Vietnam has never listed them among its recognised ethnic groups, and a Vietnamese press piece on the subject framed the whole matter as a journey toward being named. The numbers reflect the ambiguity: Ethnologue counts around two thousand speakers, while Vietnamese reporting puts the Nguồn population as high as forty thousand. The language carries the ISO code nuo whether or not the people carry a census category.

Nguồn speakers live in Laos as well, though reports of where have never quite agreed. The most specific comes from James Chamberlain, who in 1998 named a Nguồn village in central Laos: Ban Pak Phanang, in Boualapha district of Khammouane province. Boualapha is karst country, the district that holds Hin Nam No and its cave systems, pressed against the Vietnamese border where Phong Nha-Kẻ Bàng begins. The Trường Sơn crest that separates the two states is a clean line on a map and nothing at all on the ground - the same limestone, the same forest, the same families of languages spilling across in both directions. Nguồn is simply one of the things that was here before the border was.