Towards Decolonising Computational Sciences

Abstract

This article sets out our perspective on how to begin the journey of decolonising computational fi elds, such as data and cognitive sciences. We see this struggle as requiring two basic steps: a) realisation that the present-day system has inherited, and still enacts, hostile, conservative, and oppressive behaviours and principles towards women of colour; and b) rejection of the idea that centring individual people is a solution to system-level problems. The longer we ignore these two steps, the more “our” academic system maintains its toxic structure, excludes, and harms Black women and other minoritised groups. This also keeps the door open to discredited pseudoscience, like eugenics and physiognomy. We propose that grappling with our fi elds’ histories and heritage holds the key to avoiding mistakes of the past. In contrast to, for example, initiatives such as “diversity boards”, which can be harmful because they superfi cially appear reformatory but nonetheless center whiteness and maintain the status quo. Building on the work of many women of colour, we hope to advance the dialogue required to build both a grass-roots and a top-down re-imagining of computational sciences — including but not limited to psychology, neuroscience, cognitive science, computer science, data science, statistics, machine learning, and artifi cial intelligence. We aspire to progress away fromthese fi elds’ stagnant, sexist, and racist shared past into an ecosystem that welcomes and nurturesdemographically diverse researchers and ideas that critically challenge the status quo.

Algorithmic injustice: a relational ethics approach

“…When we see bias and discrimination, what we see is problems that have surfaced as a result of a field that has thoughtlessly inherited deeply rooted unjust, racist, and white supremacist histories and practices. 56 As D'Ignazio and Klein 42 contend, ''addressing bias in a dataset is a tiny technological Band-Aid for a much larger problem.'' Furthermore, underlying the idea of ''fixing'' bias is the assumption that there exists a single correct description of reality where a deviation from it has resulted in bias.…”

Reflections on Binary Sex/Gender Categorization in Magnetic Resonance Tomography and its Future Challenges

“…And on the other hand, we need to appreciate how these concepts of difference (sex, gender, class, and “race”) are intersectionally intertwined with each other. For interdisciplinary or rather intersectional approaches, we need to ask to what extent the categories of “race” or class have found their way into the statistical measurement strategies of contemporary brain research ( Abiodun 2019 ; Birhane and Guest 2020 ; Rollins 2021 ).…”

AI Content Moderation, Racism and (de)Coloniality

“…In a similar vein, Birhane and Guest (2020) argue against the tokenistic appointment of a few black men and women in positions of influence without changing and challenging the underlying structures that uphold current systems of power. They therefore argue for a gradual process of decolonization through looking at the history of the disciplines involved in the production of new technologies such as AI, and an opening up of the field to those who challenge established orthodoxies (Birhane & Guest, 2020). While Birhane and Guest propose a historically nuanced understanding, Mohamed et al (2020) look to the future, proposing the development and integration of foresight in the design of technological systems, such that it can recognize and preempt any systemic discrimination.…”

“…What may be an alternative position? In the final section of the article, following Quijano (2007a, b) and more recent works by Benjamin (2019), Birhane and Guest (2020), Mohamed et al (2020) and Adams (2021), we develop a sketch for a decolonial approach to AI in content moderation.…”