Q&A: Exploring How Refusal Can Be a Form of Design

Recently published in the ACM Journal on Responsible Computing, Jonathan Zong who is a graduate student at MIT, earned his SM ’20 degree. Co-author J. Nathan Matias, who obtained his SM ’13 and PhD ’17 from Cornell Citizens and Technology Lab, delve into the realm of data ethics to discuss the concept of refusal. In their freely accessible publication titled Data Refusal From Below: A Framework for Understanding, Evaluating, and Envisioning Refusal as Design, they propose a four-dimensional framework to illustrate how individuals can reject technology misuses, emphasizing how refusal, like design, can shape alternative futures.

Zong, a doctoral candidate in electrical engineering and computer science at MIT, also serves as the 2022-23 MIT Morningside Academy for Design Design Fellow and contributes to the MIT Visualization Group. Here are the insights he shared in a Q&A.

Q: How would you define the concept of “refusal,” and what are its origins?

A: Originating from feminist and Indigenous studies, refusal embodies the act of saying “no” without necessitating validation. Scholars like Ruha Benjamin have explored refusal in the realms of surveillance, race, and bioethics, viewing it as a complementary aspect to consent. Others, such as the creators of the “Feminist Data Manifest-No,” perceive refusal as a catalyst for envisioning improved futures.

Benjamin presents cases where certain individuals lack equal opportunities to refuse, citing instances involving genetic data and refugee screenings in the U.K. This inequality in decision-making underscores the broader essence of refusal, extending beyond rejecting specific options to challenging the entire spectrum of choices available.

Q: What sparked your interest in exploring refusal as a design concept?

A: In my study of data ethics, I’ve pondered the integration of processes in research data collection with a focus on consent and opt-out options, emphasizing individual autonomy and the provision of data usage choices to individuals. However, mere availability of choices is insufficient for data privacy, as disparities may exist in accessibility to choices or lead to dilemmas where all options are unfavorable. This realization led me to delve into the notion of refusal as a means to question the authority of data collectors and challenge their credibility.

The core premise of my work is that refusal constitutes a design activity—a purposeful endeavor to reshape our socio-technical environment by exerting influence. In a similar vein to design, refusal is creative and geared towards generating alternative potentials and futures. Viewing refusal through a design lens allowed me to establish a universal discourse on refusal and envisage unexplored forms of refusals, drawing from academic and journalistic perspectives.

Q: Why are data privacy and collection significant concerns?

A: Our paper highlights the substantial impact of utilizing data for facial recognition surveillance in the U.S. Everyday actions like social media posts or public space walks may inadvertently contribute to training facial recognition systems. For instance, a tech firm might compile photos from social media platforms to develop facial recognition algorithms sold to governmental entities. In the U.S., these systems are predominantly employed by law enforcement to monitor communities of color. Applying concepts like consent or opt-out in these scenarios proves challenging due to their prolonged nature and multi-institutional involvement. In such instances, refusal emerges as a pivotal approach, at both individual and communal levels, for affected individuals to exert influence or agency, despite lacking official channels for doing so.

Q: How do disempowered communities bear the brunt of these issues?

A: Communities impacted by technologies are frequently sidelined from the design process of said technologies. Refusal, thus, serves as a significant channel for expressing values and priorities among those excluded from early design deliberations. Initiatives against technologies like facial recognition—ranging from legal confrontations with companies to advocating for stricter regulations or direct actions like disabling security systems—may not adhere to conventional participation norms in design processes. Yet, these actions offer refusers excluded from mainstream participation avenues an opportunity to voice concerns.

Of particular interest is the movement surrounding Indigenous data sovereignty, exemplified by organizations like the First Nations Information Governance Centre, which aims to amplify Indigenous community voices in data collection, rejecting inadequate representation in official Canadian health data. This movement underscores the potential of refusal not only to reject existing norms but also to propose constructive alternatives akin to the essence of design. Refusal is not solely about negation but also about paving the way for diverse futures.

Q: Could you elaborate on the proposed design framework?

A: Refusals manifest diversely across contexts and scales, necessitating a comprehensive framework to unify seemingly disparate actions under a broader umbrella. Our framework encompasses four facets: autonomy, time, power, and cost.

Take, for instance, the case of IBM constructing a facial recognition dataset from individuals’ photos without consent. Various forms of refusal surfaced in response, exemplified by individual opt-outs, collective resistance through class-action lawsuits against IBM, and municipal legislatures passing ordinances banning governmental facial recognition utilization. Evaluating these instances through the framework underscores shared and contrasting elements. The framework accentuates diverse autonomy strategies—ranging from individual opt-outs to collective actions—and timelines, whereby opt-outs and lawsuits address past grievances while legislation anticipates averting future harms. Variations in power dynamics are observed, with individual photo withdrawals exerting minimal influence on IBM, contrasting with potential long-term impacts of legislative changes. Concerning costs, individual opt-outs seem less demanding than other approaches, which might necessitate more time and effort, weighed against potential benefits.

The framework not only aids in describing and comparing cases across these dimensions but also encourages the exploration of innovative refusal strategies. By identifying the characteristics desired in future refusals—such as collective, proactive, influential, cost-effective—we aspire to shape forthcoming approaches and alter data collector behavior. While achieving all criteria may not always be feasible, the framework provides a means to articulate our aspirational goals within this context.

Q: What influence do you envision this research having?

A: My aspiration is to broaden the scope of design participants and validate diverse actions as legitimate design inputs. Much of the existing discourse on data ethics predominantly prioritizes the viewpoints of computer scientists striving to refine systems, potentially sidelining the experiences of those for whom the systems are currently inadequate. Hence, I hope that designers and computer scientists acknowledge refusal as a legitimate design form and a wellspring of inspiration. The ongoing dialogue is crucial and should inform the design of future systems, even if conveyed through non-traditional means.

One critical aspect emphasized in the paper is that design transcends software; embracing a socio-technical standpoint, designing encompasses software, institutions, relationships, and governance structures revolving around data utilization. I advocate for individuals beyond software engineers, such as policymakers or activists, to recognize their integral role in the technology design ecosystem.