Autocorrecting for Whiteness

97 Pages Posted: 10 Mar 2021 Last revised: 12 Mar 2021

See all articles by Rashmi Dyal-Chand

Rashmi Dyal-Chand

Northeastern University - School of Law

Date Written: March 8, 2021

Abstract

Autocorrect presumes Whiteness. Across a range of products and applications, autocorrect consistently “corrects” names that do not look White or Anglo. Sometimes autocorrect changes names to their closest Anglo approximations (as in Ayaan to Susan). Sometimes it suggests replacements that are not proper names (as in DaShawn to dash away). Often, autocorrect asserts the implausibility of non-Anglo names by underlining them in red. Autocorrect’s changes to names such as these are not just trivial product glitches. In a world rife with the multiplying effects of algorithmic bias in increasingly essential domains of decision-making, autocorrect produces social and cultural harms that disproportionately affect communities of color and those who do not have Anglo identities.

Harnessing both empirical evidence and theory, this Article argues that while autocorrect’s Anglo bias harms such individuals and communities, it adds value to the intellectual property of autocorrect’s proprietors as well as to the “status property” of more privileged users. We all increasingly rely on smartphones, tablets, word processors, and apps that use autocorrect. Yet autocorrect incorporates a set of defaults—including dictionaries—that help some of its users to communicate seamlessly at the expense of others who cannot. It is a simple but powerful means of self-realization, providing a modern forum for the reinstantiation of Cheryl Harris’s concept of Whiteness as property. It is a medium for governing social relations that depends on the devaluation of non- Anglo names. It is a form of smart technology that maintains structural racism today.

The essential nature of autocorrect technology—and its far-reaching effect in structuring social, cultural, and even epistemic understandings of our world—demands legal intervention to fix autocorrect’s Anglo bias. Drawing on core norms from property law, as well as consumer law and culture, this Article proposes design principles for ensuring more transparency, access, and participation in the design and deployment of autocorrect technology.

Keywords: algorithms, bias, whiteness, identity

Suggested Citation

Dyal-Chand, Rashmi, Autocorrecting for Whiteness (March 8, 2021). 101 Boston University Law Review 191 (2021), Northeastern University School of Law Research Paper No. 403-2021, Available at SSRN: https://ssrn.com/abstract=3800463

Rashmi Dyal-Chand (Contact Author)

Northeastern University - School of Law ( email )

416 Huntington Avenue
Boston, MA 02115
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
180
Abstract Views
1,601
Rank
343,502
PlumX Metrics