Published June 2, 2026 | Version v1
Preprint Restricted

Reading Mediated Text Objects: Evidence Profiles and Interpretive Ceilings in Digital Humanities

Authors/Creators

Description

Digital humanities increasingly reads texts after they have passed through translation, edition
making, platform description, archival processing, and extraction. This article asks a field-level
question about that condition: how much interpretation can such mediated text objects still bear,
and where must a computational reading stop? It answers through two main cases read through the
same surface-oriented arrangement: Hamlet as an English play mediated through Guizot’s French
and Wieland’s German strata, and Dante’s Inferno as an Italian canto sequence mediated through
English extraction. It does not propose a new metric or evaluate translation quality; instead, it
treats computational outputs as evidence whose interpretive warrant depends on the mediated ob
ject that produced them. Hamlet preserves a near-uniform language-independent surface profile
across EN–FR and EN–DE pairs while changing scalar magnitude, especially in Guizot’s Acts III
and V, whereas Dante shows canto-level surface variation, extraction imbalance, and limited clus
tering support, so its evidence stops earlier. A same-language Hamlet edition diagnostic strengthens
the reading that the Hamlet pattern is not merely a measurement ceiling. What emerges is not a uni
versal account of translation or textuality, but a way to see why different mediated objects permit
different kinds of reading.

Files

Restricted

The record is publicly accessible, but files are restricted. <a href="https://zenodo.org/account/settings/login?next=https://zenodo.org/records/20507343">Log in</a> to check if you have access.