*pu

This morpheme appears in cognate object constructions, e.g. Bengni pɯ-pɯ pɯ ‘lay an egg’, where the last element is the verb. The noun looks reduplicated in Bengni, but Apatani has pa-pu ‘egg’, with apparently distinct components. Similar cognate objects are characteristic of #301 PTB *ʔu EGG / BIRD, e.g. Lahu ɔ̀-u u ve ‘lay an egg’, where the noun appears with a prefix.

This root may well be related to a morpheme with a more general meaning of BALL / EGG / ROUND OBJECT (#1275 PTB *pu BALL / EGG / ROUND OBJECT).

孵 OC *pʼi̯ôg ‘to hatch (eggs), incubate’ ZJH; not in GSR #1233; Li 1971: (*phjəgw); Baxter 1992: (*ph(r)ju); Mand. fū.

The earliest attested use of this Chinese character seems to be the Han Dynasty work Fangyan. However, it is clearly a later graphic variant of 孚, Mand. fú, which is attested writing ‘hatch’. Thus although the character may be of later development, the word itself obviously existed at the Old Chinese time period.

Characters with the 孚 phonetic are generally placed in the OC 幽 Yōu rhyme group. However, some members of this phonetic series are found in the Middle Chinese 虞 Yú rhyme, which is not regularly derivable from the OC 幽 Yōu group.

Neither Li nor Baxter specifically discusses the difficulties of reconstructing this set of characters. Baxter (1992:757) does however list a reconstruction for 孚 (square brackets represent irregular development): 孚 [fú] < [phju] < *ph(r)ju. It is treated as an irregular development from the OC 幽 Yōu group into the MC 虞 Yú rhyme. I have therefore provided a parallel reconstruction for 孵 in the systems of Li and Baxter.

Karlgren reconstructs other characters in this phonetic series with *-ug (equivalent to the OC 侯 Hóu rhyme group), which yields a regular development into MC. However, I have provided a Karlgren-system reconstruction based on the assumption that the word belongs in the 幽 Yōu rhyme group.

The OC-PTB correspondence of finals is regular. PTB *-u and *-əw phonemically differ only in length, and show identical OC correspondences. In open syllables the long vowel *-əw (which could be written /uw/ or /uː/) is more common, and so cognate sets exemplifying this correspondence usually involve that vowel. Examples include 九 ‘nine’ (TB *d-kəw, OC *kjəgw; Mand. jiǔ), 鳩 ‘dove/pigeon’ (TB *khəw, OC *kjəgw; Mand. jiū) and #1733 PTB *r/m-bu ⪤ *pru NEST / WOMB / PLACENTA (elsewhere in this volume). It is interesting to note that this correspondence seems to be attested only after grave initials.

As for the mismatch in aspiration of the OC and PTB initials, this raises the broader issue of voicing and aspiration within and across Chinese and TB. PTB is reconstructed with a two-way voicing contrast (e.g. *p vs. *b), while Old Chinese is reconstructed with a three-way voicing and aspiration contrast (e.g. *p vs. *ph vs. *b). Voicing and aspiration correspondences between cognates are notoriously imprecise. This is because of various complex morphological processes, not yet entirely understood, at play in word families on both sides, which can affect voicing and aspiration. On the Chinese side, it has become increasingly clear in recent decades that prefixal elements, such as *s- and various nasals, can voice or devoice root initials. (See for example Baxter and Sagart 1998, Sagart 1999, and Gong 2000.) It has also been argued that Chinese aspiration is mostly, or entirely, a secondary feature. (See Schuessler 2007:58ff for a recent articulation of this view.) Similar processes have been observed in various TB languages and posited for PTB. Because not all of these processes are fully understood, and because of the complex history of individual words and word families, it is not always possible to be sure that one is comparing etymological roots, rather than derived forms, in established OC/PTB cognate sets.

For this reason the correspondence of PTB *p- with OC *ph-, for the comparison under discussion here, must be considered regular, with the assumption that aspiration in the Chinese form is a secondary development. Similarly, mismatches in voicing or aspiration will not be considered impediments in the proposal of Chinese comparanda for PTB etyma elsewhere in this volume. We assume, ultimately, that PTB voiceless initials correspond to Chinese voiceless initials, and that PTB voiced initials correspond to Chinese voiced initials, and that as our understanding of morphological processes on each side improves, these patterns of correlation will become more evident.
[ZJH]