#377 PTB *lak ⪤ *C-yak ARM / HAND / WING

This etymon, far and away the most widely attested root in this semantic area in TB as a whole, is characterized by considerable morphophonemic complexity. The basic allofam is #695 PTB *lak (1a), but in many daughter languages the lateral has become a yod, yielding forms like #696 PTB *yak (1b). By a further sound change, this yod has developed into z‑ or ʑ‑, but reflexes with these initials are also assigned to (1b). In other languages we find forms with d‑ rather than l‑. Two explanations are possible here: (a) These languages could have added a prefix d‑ to the yodated variant, yielding #699 PTB *d‑yak (1c), which sometimes developed into reflexes with voiced affricate initials like dʑ‑; in other cases, the prefix “preempted” the initial entirely, giving forms like #700 PTB *dak (1d). (b) Alternatively, we might assume that these latter forms arose directly from #695 PTB *lak via a sporadic “deltacism”, perhaps via an intermediate **lyak. A handful of reflexes contain both an l and a d, probably best explained of reprefixation after the original l had become d.¹ Other reflexes point to a prefix g‑ added to the yodated root, #702 PTB *g‑yak (1e), with secondary meanings like ‘cubit’ or ‘armpit’; some reflexes with affricated initials are assigned to this variant rather than to #699 PTB *d‑yak (see especially the Written Burmese triplet for ‘armpit’). Finally, an honorific Written Tibetan form for ‘hand’ reflects a labial prefix, #705 PTB *p‑yak (1f).

Note that all these prefixes (d‑, g‑, p‑) are deemed to have been added to the variant with y‑ rather than l‑, as reflected in the pan-allofamic formula given above, where “C-” stands for a consonantal prefix:

*d
*g ‑yak
*p

Benedict originally reconstructed this root simply as *lak, but in the revised version of STC he included prefixal g‑ in the protoform (STC #86). This reconstruction was followed by Weidert (1987:#689), and by French (1983:500) for Proto-Northern Naga. This elevation of g‑ to special proto-status seems unwarranted, however, since the d‑ prefix is actually much more widespread. There are in fact no unambiguous reflexes of a prototype *g‑lak.

It is extremely interesting to note that a number of languages have forms for FOOT that reflect the open syllables *la or *ya, both of which could also take a prefix (either d‑ or g‑ that sometimes led to affrication of the initial). This is noted with puzzlement in STC (p. 34 and n. 308), but it seems irresistible to posit some sort of connection with this etymon for HAND.