Reconstructions

Note

Also Bunun piaq ‘how many?’, Itbayaten pirah how many?, pa-pirah ‘how many (in units or number, not in time)’, Ilongot pigi ‘how many?’, Tagalog ilán ‘how many?; a few; more than two or three, but not many; several; some’, Agutaynen pirapa ‘amount; how much?, how many?’, Tiruray firoy ‘how much?, how many?, Enggano ʔa-píaha ‘how much?, how many?’, Kilivila vila ‘how many?’, Sa’a (e)-nite, Ulawa (e)-nita ‘how many?, so many; a few’, Rotuman hisi ‘how many?

Used indefinitely (several, I don’t know how many), as well as interrogatively’. The endings in several of these irregular forms are puzzling. While Bunun –q is at odds with all other languages that retain final consonants, Itbayaten –h, which could reflect PAn *S or *h, presents more subtle problems. If it reflected *S we would expect a final sibilant in all six of the Formosan witnesses cited here, and this alternative is therefore strongly disfavored. However, if Itbayaten pirah reflects *pijah the only witness that can be cited as a check is Amis. Nonetheless, since Amis pina contradicts the reconstruction of *-h, it seems best to posit PAn *pija, and assume that both Bunun piaq and Itbayaten pirah have added a final consonant through some unknown historically secondary process. Tagalog ilán for expected **pilá, on the other hand, is doubly irregular in showing loss of *p- and addition of –n, but the highly specific semantics make it very likely that this form is also a reflex of *pija.

This form is well-know as an interrogative of quantity, in which sense it has been reconstructed by Dempwolff (1938), and Wolff (2010:933). However, cognates distributed from Atayal to Maori show clearly that it also functioned as an adjective of indefinite quantity, translatable into English as ‘some, several, a few’.