Reading for Laughter: A Mixed-Methods Investigation of Humor in the Christian Bible

This dissertation investigates a phenomenon that readers have long sensed but scholarship has rarely measured: the presence of humor in the Christian Bible. Although a scattered tradition of literary study has argued that irony, wordplay, satire, and comic reversal pervade both Testaments, the field has lacked a shared taxonomy, a reproducible discovery procedure, and any formal estimate of how reliably readers can agree on what counts as humor. I address these gaps through a sequential mixed-methods design that braids together three complementary strands of evidence. First, a literary-critical close reading of the secondary literature yields a corpus of 85 candidate humor passages drawn from 15 biblical books and classified with an eight-category codebook spanning wordplay or paronomasia, hyperbole, satire or parody, verbal irony, situational irony, dramatic irony, absurdity or the grotesque, and comic reversal. The distribution across categories is far from uniform (χ² = 49.56, df = 7, p < .0001, Cramér’s V = .764), indicating that certain comic mechanisms dominate the canon. Second, a corpus-based computational strand runs lemma-based root-repetition procedures over the BHSA Hebrew and MorphGNT Greek corpora, surfacing 3,037 unfiltered candidate repetitions from 17,176 Hebrew words. Third, a simulated annotation exercise (Monte Carlo, seed 42) models three annotators across the 85 passages and returns modest chance-corrected agreement for humor type (Fleiss’s κ = .478; Krippendorff’s α = .413) alongside a striking split between low presence agreement by kappa (κ = .008) and high agreement by prevalence-adjusted AC1 (.876). Reading these results together produces the study’s central theoretical contribution: an observable–interpretive continuum that explains why humor identification is simultaneously easy and hard, and why simple reliability coefficients mislead when comic signals are rare and interpretive. A translation-comparison analysis shows how fragile these signals are, with the Isaac laughter pun preserved in only 29.7% of surveyed English renderings and the Eglon episode in 20.0%, while cross-strand convergence reaches just 7.9%. Adopting a critical-realist stance, I argue that biblical humor is a real feature of the text that our instruments only partially detect, and that honest measurement of that partiality is itself a scholarly gain.