Measuring intelligibility in signed languages

ABSTRACT

Intelligibility of spoken languages is a widely discussed construct; however, intelligibility, as it pertains to signed languages, has rarely been considered. This study provides an initial investigation of the construct of intelligibility in American Sign Language (ASL) and evaluates potential measures for self-report and expert ratings of sign intelligibility that examined the frequency of understanding, amount of understanding, and ease of understanding. Participants were 66 college students (42 Deaf, 24 hearing) who had self-rated ASL skills ranging from poor to excellent. Participants rated their own intelligibility in ASL and then provided a signed language sample through a picture description task. Language samples were reviewed by an expert rater and measures of intelligibility were completed. Results indicated that expert ratings of sign intelligibility across all measures were significantly and positively correlated. Understanding of the signer was predicted by the amount of understanding, frequency of understanding, and ASL production skills, while understanding the picture being described was predicted by ease of understanding and ASL grammar skills. Self- and expert ratings of sign intelligibility using the ASL version of the Intelligibility in Context Scale were not significantly different. Self-report of sign intelligibility for viewers of different familiarity using the ICS-ASL was found not to be feasible due to many participants not being in contact with ASL users in the relationships defined by the measure. In conclusion, this preliminary investigation suggests that sign intelligibility is a construct worthy of further investigation.

KEYWORDS:

• Sign language
• sign intelligibility
• speech intelligibility
• Intelligibility in Context Scale

Acknowledgments

The authors wish to thank Carol Convertino, Georgianna Borgna, and Dawn Walton for their assistance with the development of measures and rating of the data.

Disclosure Statement

The authors declare that they have no conflict of interest.

Supplementary material

Supplemental data for this article can be accessed here

Additional information

Funding

This research was supported by the Grant Development Funding Scheme from the Faculty of Arts and Education, Charles Sturt University. The first author was supported by an Australian-American Postdoctoral Fulbright Scholarship.