The Tai–Kadai languages, also known as Kra–Dai, Daic, and Kadai, are a language family of highly tonal languages found in southern China, northeast India and Southeast Asia. They include Thai and Lao, the national languages of Thailand and Laos respectively. Around 93 million people speak Tai-Kadai languages, 60% of whom speak Thai. Ethnologue lists 95 languages in this family, with 62 of these being in the Tai branch.
The diversity of the Tai–Kadai languages in southern China, especially in Guizhou and Hainan, suggests that this is close to their homeland. The Tai branch moved south into Southeast Asia only about a thousand years ago, founding the nations that later became Thailand and Laos.
The name "Tai–Kadai" is controversial, and arguments have been made that it should be replaced. The name comes from an obsolete bifurcation of the family into two branches, Tai and Kadai (all else). Yet the name Kadai suggests that it includes Tai, and as such is sometimes used to refer to the entire family. On the other hand, some references restrict the usage of "Kadai" to the Kra branch of the family, for which the name Kra suffices. The replacement name Kra–Dai has been suggested, as Kra and Dai are two large and well-established subgroups that are on different sides of a major historical split. The name Kra–Dai has since been adopted in several major scholarly works on the family.
Tai–Kadai consists of five well established branches, Hlai, Kra, Kam–Sui, Tai, and the Ong Be language:
In 1942, Paul K. Benedict placed three Kra languages (Gelao, Laqua and Lachi) together with Hlai in a group for which he coined the name "Kadai", from ka meaning "person" in Gelao and Laqua, and Dai, a form of a Hlai autonym. He further proposed a genetic grouping of Tai, Kadai and Malayo-Polynesian.
This classification is used by Ethnologue, though by 2009 Lakkja was made a third branch of Kam–Tai and Biao was moved into Kam–Sui.
Based on the large amount of vocabulary they share, the Kam–Sui, Be, and Tai branches are often classified together. (See Kam–Tai.) However, Weera Ostapirat believes this is negative evidence, possibly due to lexical replacement in the other branches. He also claims that morphological similarities suggest instead that Kra and Kam–Sui be grouped together as Northern Kra–Dai on the one hand, and Hlai with Tai as Southern Kra–Dai on the other. The position of Ong Be in Ostapirat's proposal is undetermined. Note that Ostapirat prefers to use the name Kra–Dai to Tai–Kadai, which he argues to be obsolete.
Norquest (2007) accepts this distinction, and adds the difficult Lakkja and Ong Be in his classification. However, he states that Lakkja may turn out to be Kam–Sui, and Be may be Tai, specifically one of the Northern Tai languages but divergent due to contact with other languages on Hainan. Following Ostapirat, Norquest adopts the name Kra–Dai for the family as a whole.
The Tai–Kadai languages were formerly considered to be part of the Sino-Tibetan family, but outside China they are now classified as an independent family. They contain large numbers of words that are similar in Sino-Tibetan languages. However, these are seldom found in all branches of the family, and do not include basic vocabulary, indicating that they are old loan words.
Several Western scholars have presented suggestive evidence that Tai–Kadai is related to or a branch of the Austronesian language family. There are a number of possible cognates in the core vocabulary. Among proponents, there is yet no agreement as to whether they are a sister group to Austronesian in a family called Austro-Tai, a backmigration from Taiwan to the mainland, or a later migration from the Philippines to Hainan during the Austronesian expansion.
The Austric proposal suggests a link between Austronesian and the Austroasiatic languages. Echoing part of Benedict's conception of Austric, who added Tai–Kadai-Japanese and Hmong–Mien to the proposal, Kosaka (2002) argued specifically for a Miao–Dai family.
In China, they are called Zhuang–Dong languages and are generally considered to be related to Sino-Tibetan languages along with the Miao–Yao languages. It is still a matter of discussion among Chinese scholars whether Kra languages such as Gelao, Qabiao, and Lachi can be included in Zhuang–Dong, since they lack the Sino-Tibetan similarities that are used to include other Zhuang–Dong languages in Sino-Tibetan.
Vovin (2014) proposed that the location of the Japonic Urheimat (linguistic homeland) is in Southern China. Vovin argues for typological evidence that Proto-Japanese may have been a monosyllabic, SVO syntax and isolating language, which are also characteristic of Tai-Kadai languages. The following lexical comparisons between Proto-Japonic and Proto-Tai are cited from Vovin (2014).
|Proto-Tai||Tone in proto-Tai|
|Side||*pia||H||*Ɂbaïŋ ?< OC *bʕâŋ||C1|
|Aunt||*-pa in *wo-n-pa||H||*paa 'elder sister of a parent'||C1|
|Wife, woman||*mia||L||*mia 'wife'||A2|
secondary voicing in Tai
(space & time)
|Edge||*pa, cf. also *pasi||H, HH||*faŋ
|Insert||*pak- 'wear shoes, trousers'||H||*pak||D1S|
|Mountain||*wo 'peak'||L||*buo||A2, A1 in NT|
|Split||*sak-||H||*čaak 'be separated'||D1L, š- in NT|
|Suck||*sup-||H||*ču[u]p onomatopoetic?||D1S/L, š- in NT|
|Get soaked||*sim-||H||*čim 'dip into' ?< Chin.||B1, C1, š- in NT|
|Slander||*sə/o-sir- cf. nono-sir-||H/L?, but
|*sɔɔ 'slander, indicate'||A1|
|Cold||*sam-pu- cf. sam-as- 'cool it',
samë- 'get cool'
|L||NT *ǯam > šam||C2|
but proto-Kam-Sui *to,
pace Thurgood's *tu (1988:211)
|Wing||*pa > Old Japanese pa 'wing, feather'||H||proto-Kam-Sui *pwa||C1|
|Inside||*naka < *na-ka 'inside-place'||LH||proto-Tai *ʔd-naï||SW, Sukhothai A2,
CT, NT A1
|snow||nai A2 (Shan)||ntai44||N/A|
|frost||miai A1 (Lao)||mplai44||N/A|