有關彝語支的CVQ 音節在歷史音韻學上的聲調發展，Matisoff 曾以假設的有聲前綴子音系列對此問題發表過論點，並以此假設來解釋sVQ 音節的聲調發展。本文作者認為，前者事實上可以簡單化，並以類型學上常見的解釋來取代舊的學說，詳見田雅客(2005) 。而解釋sVQ 音節聲調發展也同樣可以用更簡單的辦法來取代。本文提出的新論點可以進一步解釋、RVQ 音節。此外，其他幾種音節的歷史發展，不管是開頭聲母或前綴，也同樣可利用s-的聲調發展來解釋。所以本文提出的論點比Matisoff 的方法更簡單，應用的範圍也更為廣大。

Matisoff hypothesized a system of voiced initials governing register shift in certain Yipoic (Loloish) syllables of the CVQ-type. Along the lines of my 2005 paper -where 1 argued such a hypothesized system is not needed to explain the reflexes seen in modern Yipoic languages-here, 1 argue that such a system is not necessary to account for register shift seen in some Yipoic words of the type sVQ, either. This paper suggests that such register shifts are due to the initial s- itself, and that this effect is also seen in the 'sRVQ句Tpe syllable and even in some other syllable types. That all these shifts can be connected to the influence of initial s- (or a similar voice- less fricative) is an attractive simplification of the historical phonology of the Yipoic group

彝語支；歷史音韻學；調階變換

Loloish (Yipoic) languages, historical phonology, register shifts