上海手语的底层手型与特征赋值

The Underlying Handshapes and Their Feature Specification of Shanghai Sign Language

手语中最小独立表意单位是手势。像有声语言一样,最小独立表意单位手势也由更小的无意义成分构成,这些无意义的成分包括手型、掌向、位置和运动。在这些构成手势的无意义成分中,类似有声语的音段一样结构单位的是手型。因此,手语中的手型也叫"手型音段"。研究一种手语音系的最基础、最重要的任务就是确定一种手语有多少表层手型音段,其中有多少有哪些底层手型音位。但由于手语以视角通道为载体的复杂同时性结构特点,确定手语的底层手型音位远比确定有声语言的音位要难得多。文章根据Brentari(1998)的手语音系韵律模型,以音位学理论为依据,结合上海手语实际语料,阐述了上海手语中手型特征的有标记性与无标记性,主要特征与次要特征的功能与结构,提出了符合上海手语语料的被选手指底层音系特征赋值及其结构,确立了上海手语38个底层手型"音位"。

Sign is the free minimal linguistic unit which carries meaning in sign language. Like vocal language, the free minimal meaningful unit of sign language is also composed of meaningless elements which refer to handshapes, orientations, positions and movements. Of all the four kinds of meaningless elements, handshapes are very similar to segments in vocal language in terms of structure and function. Thus, handshapes in sign language are known as handshape segments. It is the most fundamental task in a study of sign language phonology is to identify how many handshape phonemes and what they are among the surface handshape segments. But because of the visual structure and complexity in simultaneity, it is far more difficult to identify handshape phonemes of a sign language than to identify phonemes of a vocal language. Based on Brentari＇s prosodic model of sign language phonology, this paper makes a systematic analysis of the handshape phonology of Shanghai Sign Language （SSL） in light of phonemic theory. According to the phonological markedness theory, this paper examines markedness and unmarkedness of phonological features of handshapes in SSL and holds that in SSL the joint selection features of [ stacked], [ crossed] and [ closed] are too marked to be specified underlyingly and that the feature [ spread] is a default feature so that it is unnecessarily specified underlyingly, which is well attested by the SSL data. As segmental features in vocal language can be categor/zed into primary features and secondary feature, the former play a more important role than the latter in perception. The same is true with segmental features of selected fingers in SSL, where the features of [ opposed] and [ non-opposed] under the thumb are not salient for perception and are always neglected in ＂speech＂ so that these two features are secondary features while the selection of thumb is salient and makes a crucial difference for perception so that selection of thumb is of a primary feature [ THUMB], which suggests that the primary feature is specified underlyingly and secondary features are only specified on surface. Through this strong-grounded systematic analysis, the paper modifies Brentari＇s prosodic model of sign language phonology and proposes a modified feature geometry for underlying handshapes of SSL, in which the underlying feature specification of handshapes in SSL only includes two joint features of [ base] and [ non-basel, [ ＋THUMB] （ selection or non-selection of thumb） and the features for other selected fingers （ features for the number of other four fingers except thumb）, claiming that under this new model of underlying feature specification, there are only 38 handshape phonemes in SSL.