Harmful Terms in Computing: Towards Widespread Detection and Correction

Abstract

Modern-day software development and use is a product of decades of advancement and evolution. Over time as new technologies and concepts emerged, so did new terminology to describe and discuss them. Most terminology used in computing is harmless, however, some are rooted in historically discriminatory, and potentially harmful, terms. While the landscape of individuals who develop technology has diversified over the years, the terminology has become a normalized part of modern software development and computing jargon. Despite organizations such as the ACM raising awareness of the potential harm certain terms can do and companies like GitHub working to change the systemic use of harmful terms in computing, it is still not clear what the landscape of harmful terminology in computing really is and how we can support the widespread detection and correction of harmful terminology in computing artifacts. To this end, we conducted a review of existing work and efforts at curating, detecting, and removing harmful terminology in computing. Combining and building on these prior efforts, we produce an extensible database of what we define as harmful terminology in computing and describe an open source proof-of-concept tool for detecting and replacing harmful computing-related terminology.