Sermon Chunk Explorer

One row per church. # Prosperity: Hard / Soft (shown when the verified overlay is loaded) are the headline prosperity measure — hand-verified findings from prosperity_findings_site.csv. Most churches correctly show 0 Hard.
Prosperity words (raw) is raw word frequency (includes "seed," "harvest," "increase" from ordinary Bible passages like the Parable of the Sower). NOT a measure of prosperity teaching — it's shown muted, for context only. See the verified Hard/Soft counts instead. Other lexicon columns are density = counts per 1,000 words; conditional columns count chunks where if_then_present is true. Descriptive — read the quotes before drawing conclusions.

One row per church, one cell per week (Jan–Jun 2026, parsed from the date prefix of sermon_id). Cell color = the family of that sermon's single highest-density lexicon (per 1,000 words across its chunks). The 19 lexicons are grouped into 8 families so colors stay legible; the exact dominant lexicon and its top-3 densities are in the hover tooltip. Empty = no dated sermon that week. A small ● red / ● amber dot on a cell marks a sermon that week with a hand-verified prosperity finding (hard outranks soft); hover for the count. Descriptive only.

Ratios show whether a theme travels with its theological counterweight. Blunt lexicon proxy — read the quotes before concluding. Each pair contrasts per-1,000-word densities (the grace/sin pair uses gospel-element tags). Left bar = the emphasis; right bar = the counterweight; the ratio is left÷right (higher = more emphasis relative to counterweight). Sort by any pair.

Is a theme a pervasive habit or a single-sermon spike (a series, a guest speaker, one stump speech)? For the selected lexicon, per church: total density, how many sermons carry it, what share of all its hits come from that church's single heaviest sermon, and how many sermons it takes to reach 80% of the hits. Concentrated = one sermon dominates (heaviest ≥ 40% of hits, or ≤ 2 sermons reach 80%); Pervasive = spread across many. The sparkline shows per-sermon hits across Jan–Jun. Descriptive — go read those sermons.

Per-church averages of per-sermon style metrics from sermon_summary.csv. US grade level (Flesch-Kincaid); higher = denser language — approximate, and a few sermons with broken sentence segmentation were capped before averaging (see note under the table). we_to_you = ratio of we/us/our to you/your (higher = more collective framing). repetition = repeated-language share. Descriptive.