- Proto-Sino-Tibetan: ?
- Proto-Tibeto-Burman: *b(y)et (Matisoff, STEDT)
Mentioning the taboo word 屄 was avoided in ancient Chinese literature (rhyme books included), resulting in a paucity of attestations and an elevated difficulty in reconstructing the Old Chinese pronunciation. Using modern dialectal data, Benedict (1988, circulated manuscript at ICSTLL #21) reconstructed the Old Chinese word as /b’iĕt/ invalid IPA characters (’). However, a closed-open syllable variation should probably be posited in the proto-form: *pe and *pet, as many conservative dialects show forms without a final consonant (e.g. Amoy).
This taboo word caused many originally homophonic characters to undergo irregular sound changes, e.g. 卑, 裨, such that there was almost no other common word with the same pronunciation. The development from Middle Chinese to modern Beijing Mandarin, in particular the process of syllable coda loss, however generated an additional homophone 逼, which is now basically the only commonly used word homophonic with 屄 in Beijing Mandarin. It is therefore used as an euphemism for 屄 in circumstances where the vulgar nature of this word needs to be toned down. This taboo word is still very productive in compound formation; many neologisms in Chinese are based on this taboo word, e.g. 傻屄 ("motherfucker"), 牛屄 ("awesome, cool"), 裝屄 ("pretentious"), 苦屄 ("miserable, pathetic").
Some Yue Chinese dialects have hai-shaped words meaning "vagina", e.g. Guangzhou 閪 /hɐi⁵⁵/. These are probably Tai-Kadai loanwords, compare Proto-Tai *hiːᴬ (“vagina”) (Thai หี (hǐi), Lao ຫີ (hī)), although some Tibeto-Burman languages also have phonetic parallels.
Benedict (1991) hypothesised that this PST etymon came from the root *be ("hidden"), e.g. Tibetan སྦེད་པ (sbed pa), སྦས (sbas), སྦ (sba, “to hide, to conceal; also genitals”) (though the vowels may be difficult to reconcile); Chinese 蔽. Schuessler (2007) postulated that Chinese 屄 was formed from a root *pe meaning "to open" (e.g. 闢, 披), with a suffix *-t to denote natural objects.
Benedict (1990) hypothesised a "Proto-Austro-Kadai" binome *tu‑pi containing this etymon; the first component in that binome is ~ PST *s-tu (“vagina, vulva”). Bauer (1991), on the other hand, treated this etymon as part of a binome *dzu(k)‑byet (with PST *dzjuk ~ tsjuk (“vulva”)), and adduced parallels in Hmong-Mien languages and Chinese.
Using pVt- for words denoting "vagina, vulva" appears to be somewhat common in other languages as well; compare: Proto-Indo-European *písdeh₂ (“vulva”), *pout ("cunnus") (Russian пизда (pizda), Albanian pidhi, Scots fud, Latin praeputium, Latvian peža, Lithuanian putė), Finnish pillu, Hungarian punci, pina, Proto-Kartvelian bud-, Proto-North-Caucasian *pūṭi ~ būṭi, Tagalog puday, puke (~?), Proto-Austronesian *palaq, *bediq ~ *betiq (Malagasy fala), Korean 보지 (boji). See also Proto-Human language#Vocabulary on Wikipedia.
- Old Chinese: (probably *pe ~ *pet, see note above) ("vagina")
- Middle Chinese: /*pi/ (unrecorded in major rhyme books)
- /*b(y)etᴸ/ invalid IPA characters (ᴸ) (Matisoff, 1972)
- Burmese: စောက်ပတ် (caukpat, “vagina”) (per Matisoff, 1987)
- /*batᴸ/ invalid IPA characters (ᴸ) (Bradley, 1979)