Legibility, Legitimacy, and the Indus Script : A Discourse-Theoretical Study of AI and Knowledge Production in Digital Media

In January 2025, an Indian state leader announced a $1 million prize for the successful decipherment of the Indus script (Biswas, 2025; Pragati, 2025), reigniting public interest in one of archaeology’s most enduring mysteries. This thesis investigates how social media users engage with that mystery, an undeciphered writing system of the Indus Valley Civilization, through discourse centered on artificial intelligence, computational methods, and decipherment. Drawing on a netnographic method and grounded in a Foucauldian discourse-theoretical framework (Sjölander, 2020), the study examines how knowledge, authority, and cultural identity are performed, contested, and reimagined in digitally mediated environments. The analysis is structured around five thematic clusters: (1) epistemic skepticism about computational limits, (2) ideological and nationalist framings, (3) semiotic interpretation of Indus iconography, (4) negotiations between academic and lay knowledge, and (5) irony and ambiguity as forms of epistemic resistance. The findings show that media users actively shape the symbolic and cultural meaning of the Indus script through rhetorical strategies ranging from scholarly analogies to memes and satire. These findings contribute to Media Technology in three key ways. First, they show how Reddit’s algorithms, interface, and user practices shape what knowledge gains visibility, illustrating the link between technical design and epistemic legitimacy (Malmö University, 2025). Second, they position the Indus script as both data and discursive infrastructure, raising design-relevant questions about how platforms manage uncertainty and expertise. Third, they offer insight for ethical platform design by revealing how authority is performed and amplified online, calling for greater transparency and inclusivity in how knowledge is surfaced.