#518 PTB *m/s-la(ː)y ⪤ *s-tay NAVEL

A root *laːy is set up in STC #287 with the meaning MIDDLE/CENTER, on the basis of WB ălai and Lushai (Mizo) lai.¹ Elsewhere, STC presents two additional roots: *s‑tay ‘navel’ (STC #299), based on WT lte‑ba, Jingpho dài ~ šədāi, Garo ste; and *tay ‘self’ (STC #284), based on Jingpho dāi (also dāidāi) and Lushai (Mizo) tei. Yet Benedict himself implies by a cross-reference (p. 65) that these latter two roots are really one and the same. I wish to go a step further to claim that all three STC roots (#284, #287, #299) are co-allofamic.

The interchange of l‑ with dental stops occurs in several other TB roots, including TONGUE and ARROW (and this phenomenon is known from other language families as well, notably Indo-European, where by coincidence TONGUE is also one of the best examples (cf. Latin lingua vs. pGmc *tungōn-), alongside TEAR (cf. Latin lacrima vs. pGmc *tahr- ⪤ *tagr-), BROTHER-IN-LAW (cf. Latin lēvir vs. Lithuanian dieverìs), etc.²

There is strong evidence for both a nasal and a sibilant prefix with this root (occasionally with both together, as in the Ashö Sandoway form ă`hmlai¹ ‘navel’ < *s‑m‑laːy). However, the appearance of a cluster like tl‑ in TB words for SELF or NAVEL does not by itself indicate the co-allofamy of the stop and lateral prototypes, since tl‑ is frequently the reflex of *sl‑ (especially in Kuki-Chin and other languages of the NE Indian Areal Group), so that these forms could be referred directly to *s‑laːy. See the Lakher (Mara), Miju, Bantawa, and Kham forms below.

In forms like Ao ²tɯ ²pɯ ²la, Rengma ¹a ²bvɯ ³li, it might be plausible to treat the second syllables like dimidiations (syllabizations) of the labial prefix *bV‑ ⪤ *mV‑, since in that position many other languages have elements with unstressed vowels (transcribed pə‑, mə‑, pü‑, etc.). On the other hand, when the quality of the vowel in such a syllable is clearly [u], it seems preferable to analyze this element as descending from #164 PTB *d‑puːk BELLY / VITALS / HOLLOW OBJECT / CAVE or #2111 PTB *pwa(ŋ/n) BELLY / CENTER. This is especially clear in Himalayish: cf. the Bantawa and Kham reflexes.

Both the lateral- and stop-initialled variants of this etymon are widely attested in TB. In the following sections, the reflexes are divided between the two allofams.

The semantic scope of this etymon is already quite broad, but I would like to claim that there is also a genuine phonosemantic connection with a root meaning BELT / ZONE / WAIST (see below).

There is an interesting Dumi compound saː◦khil◦tɨm ‘navel’ composed of saː◦khil ‘viscera, guts’ < khil ‘feces’, plus tɨm ‘head, leader’: i.e. the navel is viewed in Dumi as the leader of the intestines. See van Driem 1993:415.

See HPTB *laːy ⪤ *m/s‑taːy, pp. 52, 210.