Psycholinguistic studies of sentence comprehension are usually conducted on “major” world languages, through laboratory experiments involving university students in highly industrialized societies. This project aimed to broaden the empirical base of psycholinguistic research by investigating Chamorro, an Austronesian language spoken in the Mariana Islands. We asked how speakers comprehend questions, relative clauses, and other constructions on a moment-by-moment basis, as each new word is processed. Chamorro has linguistic features that differentiate it from the languages usually studied in psycholinguistic research: sentences are verb-first, the object cannot outrank the subject in animacy, relative clauses can precede or follow the noun they modify, and a special verb agreement distinguishes wh-dependencies (questions and relative clauses) from simple sentences. The goal was to use these features to investigate from a new angle how sentence comprehension works. A second goal was to develop techniques for psycholinguistic experiments that would be effective with participants of diverse ages and backgrounds in a range of cultural settings. Finally, since Chamorro is in the beginning stages of language endangerment, a third goal was to use the experimental results to explore how well the language is being maintained across different generations. 

            During the three years of the project we conducted six large experimental studies  in the three inhabited islands of U.S. Commonwealth of the Northern Mariana Islands, involving nearly five hundred unique participants whose ages ranged from 20 to 80. These studies investigated the comprehension and production of wh-dependencies and, separately, the comprehension of reflexive dependencies. The research led to a far better understanding of general processing pressures familiar from better-studied languages, such as the language-universal tendency to parse subjects first. They also revealed that morphological information can defeat general processing pressures by pointing uniquely to one parse; however, when this morphological information is absent and more than one parse is possible, comprehension is delayed. The studies employed various techniques—self-paced listening, a variant of preferential looking, and touch-tracking on a tablet computer. Touch-tracking was the most comfortable, effective technique for participants of all ages and backgrounds, and could generalize well to other small-language communities. Touch-tracking generated a rich data set, including choice data as well as dynamic timing data. Finally, although older Chamorros strongly feel the decline of language fluency in younger generations, the results showed few differences by age group. The data suggested instead that participants of all ages had full control of Chamorro morphology, but younger participants had more difficulty comprehending sentences in which multiple words had to be integrated into a coherent larger structure.

            The project had significant broader impacts. Hundreds of individuals from the islands of Saipan, Tinian, and Rota participated in the experiments. The experimental results, and what they revealed about language variation, language maintenance, and cognition, were reported in five public presentations for the community. The illustrations used in the experiments were made available for Chamorro teachers to use in their classrooms. Finally, the project involved two young Chamorros in experimental research, and provided training that contributed to the successful placement of several graduate students in industry.

Last Modified: 12/07/2016
Modified by: Matthew Wagers