RR: EAGER: Data Science Literacy for All of Linguistics

University Of Hawaii, Honolulu

Investigators

Abstract

Like other social and behavioral sciences, linguistic science is inherently data-driven, and the proper care for that data is essential if linguistics is to be a robust, reliable, and reproducible endeavor well into the future. However, many linguists today are still unfamiliar with contemporary practices for handling digital data, and recent work has identified an urgent need for immediate education across most subfields of linguistics about standards and tools for collecting, structuring, archiving, sharing, citing and evaluating linguistic data sets. While some linguistics subfields (language documentation, computational linguistics) have developed strong methods for data handling, outreach to the rest of the discipline has been deficient. This project will radically and rapidly increase the literacy of linguistic scientists at all professional levels in the management of linguistic data, from undergraduate education, to graduate, early- and mid-career training and beyond. This project aims to foster sociological change across the entire discipline of linguistics, and to bring the value of data-handling skills to the forefront of linguistics education, enabling workforce development and potential employment opportunities. Broader impacts include research opportunities in the data sciences for a post-doctoral scholar, downloadable materials with guidelines and metrics for evaluating the scholarship of language data sets for hiring, tenure and promotion, and the development and delivery of formal and informal educational modules on linguistic data management. This project is designed to enable linguistics, as a data-driven social science in which inferences about human cognition and social structure are drawn from observations of behavior, is well positioned to benefit from principles of reproducible research. Currently, there is a disparity between how much digital data is produced by scientists and how much of that data has actually been deposited or made accessible through sustainable repositories or other means. The team will reduce this disparity by increasing knowledge and resources within the language sciences as part of efforts to change the discipline's culture by reducing structural and knowledge barriers. These efforts include increasing resources and rewards (such as jobs, tenure, and promotion) for accessible data management practices. Project activities include the rapid development of educational modules at several levels: a massive open online course (MOOC) aimed at undergraduate linguistics majors, and workshops for graduate students and faculty delivered at several widely-attended professional meetings over two years. This will provide a number of offerings and educational modules designed to foster best practices in data handling and data science for reproducible research in linguistics, targeting both junior scholars and mid-career faculty.The project will also disseminate training materials widely throughout the academic linguistic community, through an open-access handbook, online formats, and conference formats. This project takes seriously the contribution of data work as an intellectual achievement in its own right, and will promote methods for increasing the ability and willingness of linguists to effectively create, manage, preserve, curate, and share linguistic data.