Eluw Ndaan Kari Seediq 賽德克語的歷史發展 The Historical Development of the Seediq Language

Abstract

This thesis reconstructs the phonology, morphosyntax, and lexicon of Proto-Seediq by applying the Comparative Method to data from five Seediq dialects (Tgdaya, Central Toda, Eastern Toda, Central Truku, and Eastern Truku) and proposes a refined subgrouping hypothesis for these varieties. The reconstructed Proto-Seediq phonemic inventory comprises 18 consonants (*p, *t, *k, *q, *b, *d, *c, *s, *x, *h, *g, *m, *n, *ŋ, *l, *r, *w, *y), 4 monophthongs (*a, *i, *u, *ə), and 3 diphthongs (*ay, *aw, *uy). Proto-Seediq had strict phonotactic constraints at word-final positions, which led to a rich array of morphophonological alternations. In the morphosyntactic domain, Proto-Seediq distinguishes between a nominative marker *ka and a genitive marker *na. The language originally had five sets of personal pronouns plus an additional clitic set that combined the nominative and genitive functions. Moreover, A detailed reconstruction of the Proto-Seediq voice system reveals four voices and three distinct sets of voice affixes. Lexical reconstruction identifies several items that require support from Proto-Austronesian or Proto-Atayal. A significant discovery is the occurrence of "special doublets"—synonymous forms with similar phonological shapes—that appear to be associated with the Atayal Gender Register System. The Seediq dialects are classified into two main branches. The Tgdaya dialect constitutes an independent branch, while the remaining dialects form a Toda-Truku subgroup characterized by shared innovations involving historical *ə, sporadic sound changes, the loss of the genitive marker, and specific lexical shifts. Within Toda-Truku, the two Toda dialects merge *g and *w, while the two Truku dialects merge *c and *s and palatalize *t and *d.