Nungish languages

The Nung or Nungish languages are a poorly described family of uncertain affiliation within the Sino-Tibetan languages spoken in Yunnan, China and Burma. They include:
 * Derung (Trung, Dulong, Drung, Tvrung)
 * Rawang (Răwang, Rvwang)
 * Nung (Anong, Along, Anung)

The Chinese name Ālóng 阿龙, sometimes misread Ayi, refers to Nung (Anong). Two other languages were formerly included under Nungish in the Ethnologue, namely Nor(r)a and Lama; however, they have recently been removed, as Nora is another name for the moribund Khamyang Tai language of Assam, and Lama (or Laemae) is a northern Bai variety that has been subsumed into the Lisu ethnic group in China.

History of classification
Grierson (1928:24) tentatively put Nung (referring to the whole Nungish family, based on what was probably a Waqdamkong Rawang wordlist from J.T.O. Barnard) in the Lolo subgroup of Lolo-Mos'o, remarking, "The language appears to form a bridge between Lolo and Kachin".

Luo (2000:325 [1954]) placed Gongshan Qiu (Dulongjiang Dulong) and Gongshan Nu (Nujiang Dulong) in the Tibetan language branch (along with Tibetan, Jiarong, Qiang, and Xibo), but also stated that the person-marking in Qiu and Nu resembles that of languages in Nepal, and suggested that Qiu and Nu might form their own separate branch. Sun (1982:2) postulated a close relationship between Dulong, Jingpho, and Deng; elsewhere (2007:567) he limits this to Dulong and Jingpho. In a more extensive passage (1983:234-247), he still maintains that Dulong and Deng should be included in the Jingpho branch (1983:243), but also concludes that based on the unique characteristics of Dulong, it arguably deserve its own branch of Sino-Tibetan, but it has more similarities with Jingpho than with any other branch (1983:247). Nishida (1987) places Dulong and Nung (a supergroup including Rawang and Anong) together into a group called Lolo-Burmese-Dulong, alongside the Loloish and Burmese branches, but places Nu (Nusu?) directly under the Burmese branch.

In her PhD dissertation, Cui Xia (2009) compares Dulong with Tibetan, Qiangic (Pumi), Burmese-Yi (Zaiwa and Hani), and Jingpho, concluding that Dulong is on a separate branch. The results pertaining to Jingpho are summarized in Dai & Cui 2009.

Matisoff (various places, e.g. 2003:692) likewise postulated a relationship between Nungish and Jingpho, and a grouping called Jingpho-Nung-Luish, but neither van Driem (2001) nor LaPolla (2003) have been able to find substantiating evidence. Thurgood (2003) and LaPolla (2003) propose that Nungish may be part of a larger "Rung" group. Matisoff (2013) now agrees that the relationship between Nungish and Jingpho-Luish is due to contact, not a close genetic relationship. He also reiterates a relatively close relationship between Nungish and Lolo-Burmese, particularly the Burmish branch (Matisoff 2013:5). DeLancey (2009) includes Nungish in the Rung group along with rGyalrong, Qiang, Primi, and Tangut, and places Rung tentatively under Burmic, on the same level as Lolo-Burmese-Naxi.

Recently, LaPolla has proposed a group of features that are characteristic of Rawang (LaPolla 2012:126), and also offered a reconstruction of person-marking in Proto-Dulong-Rawang (LaPolla 2013:470).

Scott DeLancey (2015) suggests that Nungish may be part of a wider Central Tibeto-Burman group.