Permalink : http://hdl.handle.net/2115/92165

KEYWORDS : Atayalic;Proto-Austronesian;River;Fresh water;*daNum

As noted in Ogawa and Asai (1935: 7), the forms for “river” in the Atayalic languages, ləliyuŋ in Atayal and yayuŋ in Seediq, are considered reflexes of *daNum, the form for “water” in Proto-Austronesian. This paper reevaluates this observation in terms of sound correspondences as well as historical sound changes in both Seediq and Atayal, and make the claim that the Seediq form is likely a reflex of *daNum, whereas it is gauŋ, another form for “river” in Atayal, that is the reflex of *daNum. Proto-Seediq is reconstructed as *ðaðum and Proto-Atayal as *gagum. Based on these proto-forms *jajum is reconstructed to Proto-Atayalic. This form indicates that the initial and medial consonants underwent irregular sound changes, *d ＞ *j and *N ＞ *j, that were probably triggered by assimilation. In addition, Proto-Atayalic apparently underwent a semantic shift from “fresh water” to “river.” However, this paper develops a different point of view with regard to *jajum as well as its origin *daNum, arguing that it could mean both “river” and “fresh water.” If the two meanings are separatable, there must have been an independent proto-form of “river.” However, the hitherto reconstructed Proto-Austronesian form for “river,” *sapaq, is so questionable that it cannot be safely accepted. This paper proposes that Proto-Austronesian *daNum means “river” and “water,” and its meaning shifted to “river” in the Atayalic subgroup, while in all other Austronesian languages its meaning shifted to “fresh water.”