A critique of some tone universals proposed by Woo

From the introduction: In an unpublished doctoral dissertation written in 1969, Nancy Woo proposed a set of universal prosodic features. In brief, she states that a stress feature is needed (45), but not a length feature; all length is best treated as geminction (237). For tone, three features are needed: high, low, and modify (236). The feature \u27modify\u27 involves some component besides pitch, and is needed for languages with more than three pitch levels (251). All tone languages are register systems; contours are best treated as sequences of registers, each on a separate sonorant segment (141). She illustrated her thesis primarily with data from Mandarin and other Chinese dialects. Recently, Halle has shown that Woo\u27s system of features works will for describing Slavic accent (1971), and Leben has shown that the system also helps to describe Thai (1971). In this paper I attempt a critique of two of Woo\u27s proposals: the feature \u27modify\u27, and the segmental nature of tone. In support of my criticisms, I refer to published materials, mainly on Amerindian tone languages