In lieu of an abstract, here is a brief excerpt of the content:

A REFERENTIAL COMMUNICATION EXPERIMENT WITH FOREIGN SIGN LANGUAGES I. King Jordan & Robbin Battison A great deal of research has been done by psychologists and psycholinguists in the area of communication accuracy or intelligibility (cf Mehrabian & Reed 1968). Simply, intelligibility is a measure of how well a receiver can understand the communication of a sender. In order to arrive at a precise measure of intelligibility, communication researchers have made extensive use of what is known as a referential communication design. First used by Carroll (cf Osgood & Sebeok 1965: 200) and used most notably in the work of Krauss, Glucksberg, and their associates (Krauss 1968; Krauss & Glucksberg 1969; Glucksberg & Krauss 1967), a referential communication setting is a situation in which one person (S, the sender) describes a specified referent to another person (R, the receiver). Because the referent is known to the experimenter, a measure of intelligibility is very simple and straightforward. If, after attending to a sender's communication, a receiver can identify the correct referent from among others, the communication is said to be intelligible . Referential communication experiments have been used sucessfully by researchers studying the sign language communications of deaf individuals, from young, school-age children (Hoemann 1972) to linguistically mature adolescents (Jordan 1975). Most frequently, the referents which have been used with deaf subjects were either photographs or drawings. These kinds of stimuli allow for a great deal of control, while at the same time they give the communication task a lifelike aspect. Previous work (Jordan 1973) has shown that the less artificial the stimuli and task are, the more willing subjects are to "play the game" and the more confident the researcher can be that the subjects are really trying to communicate to one another. The investigation reported here was designed to compare communication accuracy (or intelligibility) within and between various sign languages from different countries. One major impetus for the study was the knowledge that while there is no natural, universal sign Jordan & Battison language, it is often reported that deaf people who do not share the same sign language can communicate with each other much more easily than two hearing people who do not speak the same language (Battison & Jordan 1976: pp 53-68 above). While this is anecdotal in nature, it does resolve itself into a readily testable question: How well do signers of one country understand signers foreign to them? In order to test this question, a referential communication setting was used. This setting, in addition to providing a great deal of control and a precise measure of intelligibility, allows for the collection of various kinds of data, including not only communication accuracy or intelligibility, but also length and content of communication. Since it is so widely held that sign language is an iconic gesture system, a careful analysis of the content of controlled sign language communications should provide some insight into the truth or falsity of such claims. Method. Subjects, both American and foreign, were all deaf individuals who had used sign language as their primary means of communication since early childhood. All American subjects were prelingually deaf, and foreign subjects were either congenitally deaf or deaf since early childhood. Foreign subjects were recruited from among visitors to Washington, D.C., during the World Federation of the Deaf Congress in July and August 1975. American signers were recruited from the Gallaudet College community and included both students and staff members. The referents for the communication experiments were 31 X 5 inch black and white photographs presented to the subjects in six 36-picture arrays. There were arrays of cars, of chairs, and of a group of three people. The car and chair arrays were made by first collecting a large group of pictures (several hundred) and then dividing this original pool into groups of similar pictures. These divisions were made by four judges (the authors and two assistants) working together, and pictures were included in a group only if all four of the judges agreed that they were similar. The judges reduced the groups to 36-picture arrays by continuing to eliminate those which were least like the others or those which had some unique, outstanding characteristic until only 36 pictures remained. These...

pdf