Reconstructions

Note

PAn *quzaN is one of the most stable and widely distributed words in Austronesian languages. Although a future verbal sense is indicated both by Ifugaw and by Hanunóo, it is likely that *q-um-uzaN functioned both as the verb ‘to rain’ and as the (subjectless) sentential comment ‘it is raining’. As is often the case in derived reduplications with only generic semantic similarity, the evidence for PMP *quzan quzan may be a product of convergence.

The most striking of the morphologically complex units built around *quzan undoubtedly is PMP *quzan batu ‘hail’, as it implies a continuous familiarity with hailstorms throughout the millennia-long migrations of the Malayo-Polynesian-speaking peoples. This is surprising, as many of the reflexes are found not in the more northerly languages of Taiwan and the Philippines , but rather in languages that are spoken very near the Equator, and in some cases close to sea level. Perhaps the agreement of Rennellese ʔua hatu with similar forms in WMP languages is also a product of convergence, since the collocation ‘stone rain’ is semantically transparent, and could easily be invented independently (cp. Tongan ʔuha maka (lit. "stone rain") = ‘hail’, where, however, maka may be a lexical replacement of earlier fatu). Finally, although the mutually corroboratory irregularity in the Batak languages and Old Javanese and Modern Javanese in reflecting /d/ rather than /j/ in this form has been noted both by van der Tuuk (1971) and by Dyen (1951), the evidence for a distinct phoneme *Z is in fact internally contradictory, and is disregarded here.