*d(w)əy

Especially in Himalayish and the NE Indian Areal Group, it is often difficult to distinguish *d(w)əy from its probable co-allofam #1181 PTB *d(i/u)l EGG / TESTICLE. Thus, Kulung wa-di (“bird + egg”) looks like Tangsa (N.Naga) wu-di, Karbi vo-ti, etc., but other Kiranti languages (e.g. Limbu, Athpare) have reflexes with -n, apparently from *-l. Kanauri, Lepcha, and Tibetan retain -l. For now, we are assigning all Himalayish reflexes of this word-family to *dil, though a better Proto-Himalayish reconstruction would be *di-l. Similarly, some languages of the NE Indian Areal Group retain overt reflexes of *-l, but many have forms with open syllables. Our assignment of some NE Indian forms to #300 PTB *d(w)əy EGG / TESTICLE rather than #1181 PTB *d(i/u)l EGG / TESTICLE remains arbitrary.

Benedict 1939:225 (“Semantic differentiation in Indo-Chinese”) analyzed compounds like Lushai (Mizo) ar-tui ‘egg’ as “bird + water”. STC postulates a connection between *twəy ‘egg’ (STC #168) and a general root *ti(y) ‘water; moist’ (STC #55 and pp. 45, 135, 196). (This latter root should actually be set up as #298 PTB *m-t(w)əy-n ⪤ *m-ti-s WATER / FLUID / LIQUID / RIVER / SOAK / SPIT.¹ The proposed connection between EGG and WATER is complicated by the related forms for EGG with final *-l. STC (n. 149) admits that Dhimal has different forms for EGG (tui) and WATER (tśi), and yet a third form in hna-thi ‘snot’. Elsewhere in STC (p. 135, discussing the Karen cognates; and p. 196, in connection with the putative two-tone contrast for PTB), Benedict suggests that the etyma for EGG and WATER do indeed descend from separate allofams, different in both tone and initial at the PTB stage.

There is a further semantic connection between EGG and HEEL (both being smooth and rounded). Cf. Lushai (Mizo) ar-tui ‘egg’ (perhaps “bird-water”, ke-ar-tui ‘heel’ (“foot-egg”, i.e. “foot-bird-water”). The syllable -ar- must be bleached of all avian meaning by the time it gets incorporated into HEEL. See Matisoff 1994 (“How ‘dull’ can you get?”), which also brings BUTTOCK into the same network of semantic associations as HEEL.

The putative semantic connection between BIRD and WATER is strengthened by compounds of BIRD with other roots for WATER, e.g. < #2080 PTB *r/s/g-wa WATER / RAIN (Tangkhul tara ‘water’, har-ra ‘egg’ (har ‘bird’), hai-ra ‘semen’).² Less clear, but possibly a parallel formation is Muya vɑ³³ vɑ⁵⁵ ‘egg’ (vɑ³³ ‘bird’). Cf. also Maring wa-yui ‘egg’, with the second element < #1013 PTB *rəy WATER / LIQUID / BODILY FLUID.

This root is widely distributed, appearing in the NE Indian Areal Group, Himalayish, Jingpho-Asakian, Karen, and Qiangic (including Tangut), and perhaps in a few Loloish forms (Xide, Nosu) where the initial is palatalized to an affricate.

Seven forms from four Luish [Asakian] languages (Sak Bawtala, Sak Dodem, Ganan, Kadu) cited in Luce 1985 (Chart L) have the confusing gloss ‘Penis/Testicles’. Since there are two phonologically similar etyma in this area, *ti EGG/TESTICLE and #641 PTB *ti-k PENIS, these forms have presented serious problems of analysis, and have been reglossed to mean either ‘penis’ or ‘testicle’, but not both. See Etymon Note under #641 PTB *ti-k PENIS, below.

See HPTB *twəy ⪤ *dwəy ‘water; egg’, p. 195.

蚳 OC *dʼi̯ər GSR #590m, ‘ant egg’; Li 1971: (*drjid); Baxter 1992: (*drjɨj); Mand. chí.

This rare character does not appear in Li or Baxter. However, reconstruction in either system is not in doubt based on reconstruction of other characters with the same phonetic and identical Middle Chinese pronunciation (e.g. Baxter 1992:750 坻 < drij < *drjɨj (Mand. chí)). But it is worth noting that at least one character with the same phonetic is placed in the 微 Wēi group by Baxter and reconstructed with *-ij: 鴟 tsyhij < *thjij (Mand. chī), presumably because it is found rhyming with a 微 Wēi group word (Shijing 24.3A). If this character is not simply an exception, then according to Baxter’s rhyme group division, characters with the same phonetic are spread across the two rhyme groups. In the absence of rhyming evidence, *drjij is also a possible reconstruction for 蚳 in Baxter’s system.

Assuming the semantics are not problematic,³ the correspondences are quite sound. Examples of OC *-id (Li 1971) corresponding to TB *-əy are numerous, and include 死 OC *sjid ‘die’ (Li), TB *səy, Mand. sǐ; 四 OC *sjid ‘four’ (Li), TB *b-ləy, Mand. sì; 屎 OC *hrjid ‘excrement’, TB *kləy, Mand. shǐ. The TB final *-i should not be troubling; we would expect this relatively rare PTB final to correspond to OC in the same way as *-əy (in parallel to TB *-əw and *-u; see the discussion of #1168 PTB *pʷu EGG / BIRD / ROUND OBJECT). On the correspondence of OC medial *-r-, see the discussion of 中 under #3584 PTB *(t/d)uŋ ⪤ *ts(y)uːŋ MIDDLE / CENTER / INSIDE / NAVEL.
[ZJH]