SignType: A cross-linguistic database of signs

University Of Connecticut, Storrs CT

Investigators

Abstract

Hundreds of sign languages have emerged independently everywhere that deaf people had the opportunity to form cultural communities. How similar are these sign languages and in what ways do they differ? These are questions that linguists try to answer. This project aims to develop a tool to compare one aspect of sign languages, namely the visual shape of signs. With NSF support, and the help of an international panel of sign language researchers and database experts, Drs. Harry van der Hulst and Rachel Channon will build SignType, a database of the visual shape of signs. This database will be a web-based resource for the linguistic research community and the deaf community. SignType will include photographs and videotapes of sign inventories of a number of sign languages but the most important part of this database will be an analytic data structure for each sign, using a universal coding system. The coding system will include codes for the shape, movement and location of the hands, facial expressions, and many other characteristics. SignType will combine several inventories that other sign language researchers have made by translating their separate coding systems into the universal coding system. SignType will allow researchers to answer questions like: How many different handshapes or locations are there? Are there preferences? Once developed, SignType can also provide a basis for developing a universal (graphic) notation system for signs. There are, in addition, many potential practical applications such as providing material for dictionaries or sign animation/recognition software.