03 / Methodology
A path through provenance, selection, and reading that preserves documentary states and separates observation from inference.
We identify the programme, edition, and document type, recording URL and time context. Social profiles or current biographies do not fill gaps.
A complete programme, partial source, present bio field, comparable text, inaccessible resource, and pending search are different states.
The document unit is a bio attached to a published participation. Talk titles and abstracts are not treated as biographical evidence.
Evidence is a short substring of the original text in its original language. Analytical paraphrase is displayed separately.
A public bio permits a proposed textual mechanism and a recorded alternative reading or limit. It does not verify external credentials.
04 / Pilot
The documentary pilot contains 53 official biographies from 2024: 13 from Black Hat USA, 13 from DEF CON 32, 13 from 38C3, 13 from the RSAC addendum, and one from USENIX Security. Selection sought different formats and lengths; it was not probabilistic. RSAC is partial and USENIX contributes only one comparable bio in its eligible frame. The groups are not equivalent samples of their conferences.
05 / Codebook
06 / States
A ‘present’ signal requires explicit wording. ‘Absent in reviewed bio’ describes only the complete available text. Ambiguity, noncomparability, and documentary non-evaluation retain their own states; they do not become zeros. A passage may inform more than one family when the rules allow it.
07 / Inference
Vignettes were chosen contrastively after examining the corpus. They illustrate mechanisms and tensions in the instrument without estimating prevalence, saturation, or population differences. AI assistance organized reading and literal quotations; agreement is neither human validation nor a reference truth.