Doctoral Dissertation Improvement Grant: Coarticulation and the phonetics of fingerspelling

这个项目的目的是研究签名者如何产生美国手语(ASL)指纹拼写。研究手指拼写提供了发现手形的上下文和时间限制的变化的机会，这些变化在手势方面相对有限。这项工作建立在手语产生的语音系统基础上，但详细关注了美国手语中构成手形的具体方面。视频和运动捕捉数据的定量测量将用于分析ASL指纹拼写中基于上下文的手形变化。在这个过程中，将创建一个大型的指纹拼写语料库。这样的语料库将广泛用于未来的研究，并将对自动手语识别特别有用。这项工作的发现将有助于了解人类语言是如何产生的。ASL中的手形能够以一种在声道中很难研究的方式告诉我们对发音手势的理解。具体地说，各个发音器有许多不同的主动和非主动配置。这与许多口语发音装置不同，许多口语发音装置在活动时通常具有单一配置，而在非活动时通常具有另一种配置。由于手语发音器不在身体内部，因此可以比语音发音器更容易获得准确的测量结果，以分析活动和非活动时发音器的配置差异。这些发现将进一步发展跨语言变异理论，更重要的是，跨情态变异理论；语言学家可以更好地理解语言产生中的哪些属性是语言固有的，哪些属性是特定于情态的。此外，本研究还将进一步加深对手形在手语音系学中的理解。这项工作有许多更广泛的影响。首先，手语的自动识别需要手形变化的知识。这项工作为本地ASL用户建立了指纹拼写的一般标准。有了指纹拼写特定特征的量化规范，就可以开发衡量标准和测试，以确定哪些类型的产品超出了典型签名者的范围。其次，有研究表明，手指拼写能力和识字能力之间存在相关性。了解有关手指拼写产生的基本语音事实将有助于未来更详细地研究手指拼写的感知。此外，了解指纹拼写是如何产生和感知的，将使我们能够更详细地研究这种相关性。

The aim of this project is to study how signers produce American Sign Language (ASL) fingerspelling. Studying fingerspelling provides opportunities to find contextual and time conditioned variation in handshape that are relatively limited in signing. This work builds on phonological systems of sign language production, but with a detailed focus on the specific aspects that make up handshapes in ASL. Quantitative measurements of video and motion capture data will be used to analyze contextually based handshape variation in ASL fingerspelling. In this process a large fingerspelling corpus will be created. Such a corpus will be widely available for future research, and will be particularly useful for automatic sign language recognition. Findings from this work will contribute to knowledge of how human languages are produced. Handshape in ASL has the power to inform our understanding of articulatory gestures in a way that is difficult to study in the vocal tract. Specifically, individual articulators have a number of different configurations that are active and nonactive. This is unlike many spoken language articulators, which frequently have a single configuration when active, and another when nonactive. Because sign language articulators are not inside the body, precise measurements to analyze the differences between the configurations of articulators when active and nonactive can be obtained more easily than for speech articulators. These findings will further theories of variation cross-linguistically, and more importantly, cross-modally; linguists can better understand what properties in language production are inherent to language, and what properties are modality specific. Additionally, this research will further the understanding of handshape in sign phonology. This work has a number of broader impacts. First, automatic recognition of sign languages requires knowledge of handshape variation. This work establishes general norms for fingerspelling in native ASL users. Having quantitative norms of specific features of fingerspelling allows for the development of metrics and tests for what types of productions fall outside of the range of typical signers. Second, there has been research showing a correlation between fingerspelling ability and literacy. Understanding basic phonetic facts about the production of fingerspelling will allow for more detailed future work on the perception of fingerspelling. Furthermore, understanding how fingerspelling is produced and perceived will enable the study of this correlation in more detail.