Queer Machine Reading and Close Reading in Literary Studies in the Age of AI

How, in this age of artificial intelligence, do we imagine a queer machine reading where cultural specificities are (re)discovered, reclaimed, and reattributed through queer theory, close reading, and calculative practice? Heejoung Shin sets forth the groundwork to explore this recalibrated framework, paying special attention to the word embedding model both as a calculative machine learning methodology and queering strategy loaded with subversive power, a technique that tends to highlight the single and particular over the general. Shin discusses how static and contextual word embeddings can be utilized for queering projects through close reading, explaining their conceptual and empirical foundations, and providing examples demonstrated through close reading. While contextual word embeddings are known for better addressing polysemy and highlighting the functionality of key non-normative language, Shin asserts that static word embeddings might be a more satisfactory approach for close reading, in their capacity to discover related concepts. Lastly, building upon digital literary studies on the topic of distant reading, Shin stretches the disciplinary definition of close reading, situating it in the broader context of quantification, in alignment with distant reading.