Cross-media research on #MeToo with Twi-XL

This blogpost introduces the PhD research of Sarah, who is working with Twi-XL to develop cross-media research methods for societal research with Jupyter Notebooks.

Having a genuine interest in power relations and how they can be transformed by (computational) media, my case study focuses on #MeToo debates in the Netherlands and the role of and interaction between different media in shaping the discourse around topics like sexual intimidation or transgressive behavior. Having access to different data and archives such as Twitter data (Sang and Bosch, 2013), television and radio records (van der Molen et al., 2019), newspaper (ASCoR), websites (KB webarchief), Twi-XL offers a unique opportunity to study the evolution of feminist activism over time (deep), across different media outlets (broad) and within a national context (specific).

Cross-media research is a relatively new and challenging field, requiring careful reflection and design of research methods. Each media outlet or platform has its own logic of governance and institutional or corporate politics, peculiar ways of shaping and enabling discourse and mediating particular narratives. Whereas for studying digital media, the approach of ‘digital methods’ (Rogers, 2013) attends to the inbuilt ‘natively digital’ affordances of media (e.g., rankings, APIs, links) and repurposes them in critically reflected ways for societal research, for legacy media, one must instead rely on so-called ‘virtual methods’, which digitize non-natively digital content and translate it into a digital format (e.g., automatic speech transcription of television broadcasts or scanning of print media). The former approach resides in (new) media studies, whereas the latter is more common in digital humanities.

This methodological separation is not a clear-cut or simple affair since there exist overlaps, when for example legacy media offer their content ‘online’ via social media platforms, like broadcasters moving onto YouTube, or newspapers moving onto Twitter. Sometimes, online engagement can shape the narratives and format of legacy media and vice versa (Moe et al., 2016). These dynamics remain generally understudied, where Twi-XL is an attempt to demonstrate the usefulness of studying such complex inter-media relations by leveraging and connecting existing archives and datasets for societal research.

To study the role of different media outlets in the case of #MeToo is not only interesting from a media studies perspective, but at least as important from a feminist research perspective. Understanding feminism–which encompasses a vast amount of different theories, thoughts and practices–as ‘a movement to end sexism, sexist exploitation, and oppression’ (hooks, 2000), the strand of feminist media studies puts a particular focus on the role of media within this collective struggle.

Traditionally, legacy and mass media but also film and pornography were predominantly seen as antagonistic forces for (second wave) feminism in their degrading and sexist representation of women. In its relation to media, feminist practice and activism relied instead on crafted and print-based, “alternative” media, which were (and still are) circulated amongst closed activist groups and feminist book shops and publishers (Eichhorn, 2010).

With the rise of the Internet and information technologies around the 1990s, more loosely united (third wave) feminists started leveraging the crafty and cybernetic potential of the web through creative and experimental manipulation with basic web technologies such as HTML and CSS, blogs, or online journals (Eichhorn, 2010). However, there was little political activity or communication across these loosely spread feminist media sites. This has fundamentally changed with the rise of social media platforms, leading to what is now anticipated as an emergent fourth wave feminism that actively exploits and repurposes the digital native affordances to render collective struggles of sexist oppression visible, and to thereby exert political power (Schuster, 2013). In that way, feminist movements such as #MeToo started hijacking the most visible and populated media platforms of today (Twitter, in the case of #MeToo) rather than taking a radically antagonistic stance towards the big tech companies operating these massive platforms. In that sense, social media platforms appear to facilitate novel and powerful grounds for feminist activisms that combine the public reach of legacy media whilst allowing for creative repurposing of their inbuilt affordances such as hashtags, which, taken together, allow for networking and spreading feminist thoughts, voices, concerns at beforehand impossible scale.

Although this trajectory suggests that feminism remains largely antagonistic or invisible regarding legacy and mass media outlets, and instead got increasingly intertwined and visible with the rise of digital platforms, there remain open questions to what extent this is the case–or if not ultimately, the relation between legacy or mass media and feminism changed alongside the rise of the “fourth wave” social media-driven feminism. This research question sits at the core of my PhD research with a focus on the Dutch #MeToo debate, and seeks to contribute to growing calls for more contextual and situated research on feminist movements like #MeToo (Mendes et al., 2018), and is as well a response to the general lack of studying cross-media relations. Ultimately, this question hopes to enlarge knowledge on the role, responsibilities and power of different media outlets for feminism, and how these carry both potentials but also issues as individual as well as entangled spheres.

The question of how media can work towards or against feminist goals and concerns is not only of interest in terms of public “consumer media”, but is of equal importance for the computational media that are used for the actual case study and research. Being confronted with a wide variety of different institutions offering valuable but also privacy protected data, and the task to channel and connect these different types of data through innovative but also different medium-specific research methods (e.g., digital methods and virtual methods) poses a crucial challenge within my PhD research. Therefore, taking again a feminist theory-informed media studies perspective, my PhD focuses on the critical reflection and design of computational methods at the level of software (Jupyter Notebooks), infrastructures (GitLab, APIs), archives and data (KB webarchief, TwiNL, Sound and Vision television archive, ASCoR newspaper data), and algorithms (NLP techniques, computer vision), amongst others. Acknowledging that each of these components shapes and conditions the possibilities for researchers to synthesize and produce cross-media narratives differently, a key question is how feminist thought can inform a computational design of software, infrastructures, archives, data and methods that is inclusive, sensitive, intersectional, and contextual, amongst others.

Given the growing concern that computational tools or methods–the technical possibilities –dominate over the actual issue (domain) at hand, and thereby risks superseding expert knowledge or lacking theoretical grounding in the field, it is crucial for my research to dive deeper into feminist theories as well as media theory. For this, I look into theories related to new materialism (Parikka, 2012), Actor-Network-Theory (Latour, 2007), or algorithm and software studies (Dourish, 2017; Rieder, 2020), as well as feminist media studies (D’Ignazio and Klein, 2016; Zoonen, 1994) and feminist writing on affect and new materialism (Ahmed, 2014; Alaimo et al., 2008). Central to my theoretical interest is finding a language (through computation and data) that allows me to speak and mediate across different media spheres whilst maintaining and respecting their peculiar and specific form, where I seek to prepare the digital (or digitized) “matter” for societal and humanities research through feminist wielding and reflection.