Defining Full Stack Feminism

Introduction

‘Full Stack Feminism’ is a call to action. It is an articulation of the need to rethink the way we design and develop our digital systems, write our code, choose our digital technologies and methodologies, and engage with “users”. It calls for a thorough engagement with diverse voices, not as virtue signalling, but as a critical and essential methodological component of research, research development and discourse. It involves reframing and reimagining the history of computing to include and highlight the vital role and contributions made by women and individuals who are not white cis men in computing and adjacent fields, like maths.

It contributes to the discourse of gender and technology, and through an intersectional, queer, anti-racist, lens challenges the heteronormative, masculinised, and neo-liberal structures which facilitate and perpetuate inequity whether manifested in binary code, or social, cultural and political ones. In short, it advocates for the need to embed intersectional feminist thought, action and praxis across our technological stacks and infrastructures.

‘Full Stack Feminism’ is, and must be, an ongoing commitment. It cannot, nor does it pretend, to resolve the known and unknown factors which create bias and discrimination in our digital technologies. The roots of digitally enhanced, programmed and algorithmic bias are deeply entangled in historic systems of oppression which continue to work against those historically excluded and marginalised in our societies. There are no quick fixes in societies, similarly, there are no quick fixes across the stacks of technologies and their infrastructures. As such, the project highlights, acknowledges, and takes account of interventions in the field, while also creating its own.

We take inspiration from intersectional feminism praxis which, stemming from black feminist thought in the 1970s and 80s, highlights the way in which systems of oppression overlap, particularly in relation to race and gender, resulting in privileges for some and marginalisation and oppression for others (see Combahee Rivers Collective, Kimberly Crenshaw,1 bell hooks,2 Audre Lorde,3 and Patricia Hill Collins.4) Intersectional feminism helps us to think beyond our own individual experience, and is a key methodological framework for this project.

The concept

‘Full Stack’ is a term borrowed from the field of computer science and software engineering. It refers to someone who can develop both frontend (user interface e.g. HMTL, CSS, JavaScript) and backend (data management e.g. PHP, MySQL, and server management). In our 2022 artist call Laurence Hill, FSF digital curator, defined full stack in terms of ‘the part that is visible to the public’ and ‘the behind-the-scenes coding’.5 This dichotomy between visible/invisible code and stacks, is relevant to our approach and utilisation of ‘full stack’ as metaphor. It helps to question and highlight the way in which software functionality and code, data processes, procedures, calculations, assumptions, statistics, profiling, collection, tracking, and surveillance are often hidden from the end-user (us!). It prompts us to think about insider knowledge versus outsider, lived experiences.

Critically, there exists related stack discourse within the humanities and Digital Humanities, including Tiziana Terranova’s (2014) ‘Red Stack Attack! Algorithms, Capital, and the Automation of the Common’,6 and Alan Liu’s (2020) ‘Toward a Diversity Stack: Digital Humanities and Diversity as Technical Problem’.7 Both Terranova and Liu reference Benjamin H. Bratton’s The Stack : On Software and Sovereignty (2015). Across this related discourse, Liu’s “diversity stack” is akin that of Full Stack Feminism, both in its articulation of the need for intersectional feminist interventions and within their discussions of diversity not as tick boxing but as a set of affirmative actions.

The concept - ‘Full Stack Feminism’ - therefore is born from these contexts actualised and fomented through the research and dialogue within the AHRC-IRC funded network IFTe (The Intersections, Feminism, Technology & Digital Humanities network, 2020-21) team.8

The Framework - Revealing Layers of Complexity

The Full Stack Feminist Framework is comprised three interconnecting, related but distinct stacks: (1) data and archives; (2) infrastructure, tools and code; and (3) access, experience, and integration. Each can be summarised as follows:

1. Data and archives focuses on the ways in which data, in all its manifest forms, is harvested, collected, described, modelled, and stored, as well as silenced, repressed, deleted, and destroyed. It is concerned with the foundations of knowledge and how certain data (or histories in archives) have been silenced and/or side-lined in favor of others.

2. Infrastructure, Tools & Code considers the underlying architecture, systems and code upon which digital projects/systems rely, asking how they are built, how bias develops through code or through the use of specific digital tools, and what, if any, exclusionary practices manifest in this space.

3. Access, Experience and Integration is concerned with the “frontend” so to speak and the way in which digital tools, resources, and ‘archives’ are accessed and experienced by users, by consumers, by archivists, and ways in which knowledge and digital cultural heritage is accessible or indeed inaccessible. It considers the developer’s point of view as well as the user/consumer’s point of view. This layer is also where certain groups are either given or denied access to infrastructures and systems, whether physical, digital or imagined.

Our stacks are not linear and do not progress necessarily in a sequence but provide a framework and focus for analysis and action - they are touchstones, references points, sites to prompt discussion and conversation. Each represents critical pressure points which we, as a research team, have identified through our experiences. The motivations and inspiration for Full Stack Feminism are multiple and varied, and are as much motivated by personal experiences - the personal is always political - as they are by high-level, “our digital technologies are riddled with problems and bias” warnings and alarms. This articulation represents our ‘feminist version of objectivity’.9 That is to say, we accept our subjectivities in these areas, and work with them rather than against them.

Our research questions, refined through collaborative discussion with the FSF research team, are as follows;

1. How has (and how can) intersectional feminist methods, praxis, and theories shaped digital arts and humanities?

2. What are the challenges and opportunities of decentering traditional voices and histories in digital arts and humanities developments?

3. How can we ethically support the needs of public digital humanities groups and organisations?

4. What tools and methods do we need to create more inclusive and socially aware technological infrastructures, systems, and code?

5. What existing intersectional feminist design principles, methods and approaches can be incorporated into digital humanities coding and software development practices?

Each question is linked to a series of actions. These include: the development of the Full Stack Feminism Toolkit; acknowledging existing and ongoing feminist histories and interventions in DH; revisiting the role of women in computing; experimenting with autonomous feminist infrastructures; among others.

Each question and their related action(s) reveal stacks of complexities, layers of entangled systems which generate new opportunities and challenges. These entanglements are explored in various elements of our work and research related to the stacks. Our aim is to be polyvocal, that is to include many perspectives and voices. This is not always easy, not always achievable - we acknowledge there will be gaps. As such, part of our research methodology included semi-structured interviews, community forums and a public survey. The results and findings of which inform our toolkit, coding workshop series, experiments on feminist autonomous infrastructures, the making of feminist objects/artefacts, and our artist in residency call and exhibition. They also inform recommendations and policy developments, which will feature in our Toolkit.

Conclusion

Simply put, full stack feminism is a concept that applies the principles of intersectional feminism to the entire "stack" of technology, from the hardware and software layers to the social and cultural aspects that technology influences and is influenced by. We see it as a way to examine the tools, systems, and codes we (re)use, as well as the patriarchal, ultra-capitalist, neoliberal, heteronormative, social, cultural and political eco-systems in which they exist. It reminds us to critique all layers of our social, cultural and technological infrastructures and to embrace intersectional feminist praxis as fundamental approach in digital humanities and related fields. In doing so, it contributes to the ever expanding field of intersectional feminist, and queer, digital humanities.10 The history and detail of which will be explored in a later publication.

Appendix - a worked through example

Implementing or applying the FSF framework can explore the following real-world scenario: a generative AI chat bot, when given specific prompts, ’demonstrates racial prejudice against speakers of African American English’.11

• For (1) data and archives, the AI model might be trained on datasets like Common Crawl, and Wikipedia, as well as social media data from platforms like Twitter and Reddit. These datasets often reflect historical biases, such as the underrepresentation of minority voices and perspectives. For example, ‘89% of U.S. Wikipedia editors identify as white’,12 which leads to a lack of diverse viewpoints and entries. Since the model is trained on internet data, it can inherit societal prejudices because online content often contains racial biases and stereotypes, which are then absorbed by the AI model.

• Regarding (2) infrastructure, tools, and code, it’s important to consider who developed the AI model. Are the developers, or company, known for prioritising ethical considerations, and have they implemented ethical guidelines and practices during the development process? What safe-guards are in place? Who is writing the code, and whose values and perspectives does the code therefore represent? How is the code reviewed and/or tested before deployment?

• Lastly, (3) access, experience, and integration, prompts us to consider what design processes underpinned the generative AI chat bot. Do we have access to an API to explore the infrastructure, or the data? What processes or mechanisms exist for user feedback? Overall, how is the user’s experience accounted for, and were users included in the development life-cycle, and at what stage? Were they from diverse backgrounds?