This project investigates the criteria through which particular objects, groups, and practices come to be considered "cults," comparing the semantic structures that emerge from scholarly literature, interviews with participants, and institutional material on sectarian drift.

Corpora

Reference lists

Interview corpus stats

26

Interviews

12002

Total words

462

Avg words / interview

5

Translated

By batch

  • Interview Batch 1: 11
  • Interview Batch 2: 2
  • Interview Batch 3: 7
  • Instagram Batch: 6

By language

  • English: 21
  • French: 5

By method

  • in person: 20
  • Instagram: 6

Full-text extraction & embedding pipeline

Separate from the bibliographic corpus above: 85 source documents have been run through extraction, annotation, and embedding, producing 44,325 embedded chunks and concepts, pooled into one 394-d shared space capturing 95.0% of variance.

85

Documents processed

44,325

Embedded points

3,000

Concept backbone

394-d

Shared space (95% variance)

Chunks embedded by corpus

  • interviews: 230 (26 docs)
  • literature: 39,236 (57 docs)
  • miviludes: 914 (2 docs)
Explore the shared embedding space (44,325 points) →