Illness narratives, or autobiographical accounts of the lived experience of pathology or disability, have been established as an effective therapeutic intervention for responding to emotional well-being related to illness (Couser, 1999, 2004, 2005, 2009; Frank; Hartman; Hawkins; Irvine & Charon; Kleinman; Mintz; Sontag). The scholarly field related to illness narratives is currently grappling with the medium’s expansion from the traditional book to digital-born narratives, however, there is limited research analyzing illness narratives built through literary hypertext. Literary hypertext is a form of digital story writing that calls on the reader to participate in the narrative’s unfolding by selecting hyperlink options which branch the narrative into nonlinear directions. There has been a revival of scholarly and public interest in literary hypertext in the past decade, owing to the genre’s culture of free production and distribution (Anthropy; Harvey). This project questions how women and nonbinary individuals with the endocrine disorder hyperandrogenism can use hypertext technology to write illness narratives that construct positive relationships between their identities and the world. Ten participants with hyperandrogenism completed a pedagogical module on building hypertext illness narratives. The corpus of this research, including participant narratives and interview transcripts, was analyzed through a feminist new materialist theoretical framework and a novel methodology called Critical Discourse Analysis for Digtial-Born Narratives. The findings of this research argue that literary hypertext technology was used by participants to visually map and manually chart experiences through the practice of hyperlinking in order to create a structure perceived as best suited for therapeutic reflection.

When Digital Humanists envision the future of Digital Humanities, Linked Open Data (LOD) often features prominently. First conceived by Tim Berners-Lee and collaborators in the World Wide Web Consortium, LOD is the connective tissue of the Semantic Web (Gracy, Maeda, Verborgh and van Hooland, Berners-Lee, Niu). Its key innovation is an ability to link data across disparate sources; its central promise is to connect data that has been hitherto siloed, and only storable and searchable in bespoke forms (W3C Consortium). In our own practice as digital humanists who lead the AdArchive Project, we have explored the potential for LOD to enrich our understanding of the complicated networks that sustained feminist periodical publishing in the 1970s. AdArchive is an LOD experiment that represents components of feminist periodicals in RDF; its goal is to build a digital archive of feminist advertisements in order to digitally represent a rich ecosystem of relationships that existed between a range of disparate entities within a larger movement (publications, publishers, editors, organizations, and authors). Using LOD, we set the groundwork for representing the connections among feminist datasets and enriching understandings of the histories of feminist cultural production.

Insofar as they both lend themselves to a decentralization of knowledge and a rejection of hierarchy, there are some clear points of harmony between feminist epistemologies and the networked logics of the Semantic Web. Nevertheless, like other classification systems, LOD is not inherently feminist. This paper argues that the feminist orientation developed in the AdArchive project can contribute to reimagining LOD’s future in digital humanities. We make this argument through a detailed description of five practical principles for authoring data in ways that align with feminist orientations. These principles include:

1) Describe everything - data democracy

2) Description is interpretation

3) Vocabularies and ontologies should not conceal data

4) Embrace messiness

5) Hold space for absence

Drawing on feminist digital humanities scholars including Catherine D’Ignacio and Lauren F. Klein, Joanna Drucker, and Lisa Gitelman and Virginia Jackson, and considering the work of Martha Nell Smith, Susan Brown, and Julia Flanders, we approach LOD data authoring as an active practice of remediation that requires careful reflexive and open-ended practice. It is vital for us to be clear that our goal is not to author feminist data, but to author data in ways that align with feminist methods, ethics, and projects. We view feminism as a practice and an orientation, not as a static quality of things, objects, artifacts, or even individual researchers. Our practice involves applying feminist principles when we author RDF triples for LOD. By outlining the principles that guide our project - and complementing them with practical descriptions of the work - we hope to provide encouragement for other researchers to participate in the praxis of imagining the equitable digital spaces of the future, digital spaces that complement humanistic and feminist thought.

This paper uses the interdisciplinary analytical methods of feminist technoscience to reveal discrepancies between Apple’s public-facing Inclusion and Diversity commitments (which report their improvement in more inclusive and diverse opportunities and corporate representations) and its design of blatantly biased software, with a focus on Apple’s AI assistant Siri. While Big Tech corporations evoke efficiency, modernity, and innovation, they also repeatedly raise concerns about exclusion and a lack of diversity through the design of products with demonstrable bias toward underrepresented groups (Noble 2018; O’Neil 2016). In particular, by mimicking and reproducing limited concepts of gender through embedded design, Siri sustains a vicious cycle of reinforcing gendered stereotypes in technoculture and to users en masse.

By combining methodologies from critical discourse analysis, critical software/code analysis, and critical design analysis, this paper draws upon the interdisciplinary approaches of feminist technoscientists who analyze artifact and platform design to reveal underlying ideologies and potential biases of technologies; specifically, we draw upon the methods of Anne Balsamo (2011), Daniela K. Rosner (2018), and Judy Wacjman (2004; 2010). Our paper contributes a comparative analysis of Apple’s proposed diversity in its workforce and its investment towards such initiatives with how Siri has evolved in design to see if these actions are mutually constitutive. In other words: are the money and efforts Apple invests into inclusion and diversity being reflected in the products they put out?

Our close reading of Apple’s Inclusion and Diversity commitments reveals the following data: from 2014 - 2021, Apple’s overall workforce improvement is a generally static pattern, with women being in the minority, and an overall 4.8% increase of women over eight years. Apple’s technology workforce had similarly minimal improvement, with men vastly dominating and a 4.4% increase of women. We compare this data with our qualitative critical code and design analysis of Siri’s available code scripts, which reveal that the software has been minorly tweaked over the years. The Anglo American software was originally voiced as female, programmed to sometimes perform “wifely” duties and even respond flirtatiously. While Apple has since removed some more flagrant traits and offensive responses, developers have not offered standardized guidelines against sexist representations (UNESCO 2019; Fan 2021), thus excusing themselves from changing other problematic design decisions unless called out by the public.

We argue that, while Apple’s more recent UX/UI design decisions reduce the immediate association of Siri’s utterances with women, Siri still presents a binarized notion of gender, showing that Apple has not done enough to mitigate their biased language models. As Big Tech companies pledge corporate social responsibility yet create technologies that diverge from these promises, this paper concludes that it is important to maintain critical discourse on harmful user effects, their mitigations, and social considerations for equitable technological development. Our identification of discrepancies among technological products’ theoretical design, company values, and product applications is ultimately in an effort to hold Big Tech companies accountable for their public-facing commitments, toward creating meaningful policy, challenging gendered systems, and creating more equitable products.