From Readers to Data : Uncertainty in Computational Literary Citizen Science

We examine uncertainty in computational literary citizen science by analysing The Hebrew Novel Project, a large-scale initiative collecting reader interpretations of Hebrew novels. While citizen science projects typically treat uncertainty as noise, we demonstrate the value of treating it as meaningful data. Through statistical-phenomenological analysis of 1,026 questionnaire responses from 349 readers, we study how readers express uncertainty, from simple question-skipping to explicit rejection of interpretive frameworks. We uncover theoretically meaningful uncertainty patterns - certain literary concepts consistently elicit more uncertainty than others, and individual readers show varying but consistent levels of epistemic humility across different aspects of literary interpretation. We argue that this "productive uncertainty" provides insight into both the nature of literary texts and the process of reading, suggesting new directions for computational literary studies that embrace interpretive ambiguity. By taking uncertainty seriously, citizen science projects can address a wider scope of interpretive phenomena while maintaining methodological rigour.