Daniel Sroka’s Second Interview with Chief Paul — Transcript Versions

How the transcript was prepared

The original rough transcript was generated using an automated live speech-to-text app. It was then corrected and revised with AI assistance, followed by manual revisions by Eric Seamons and Daniel Sroka. The versions below preserve that progression, from the original app output to the final participant-reviewed transcript.

No audio recording was available for verification during this editing process. The revised transcript is an edited reconstruction, rather than a verified verbatim record. Earlier drafts retain transcription errors, unresolved passages, and review notes that were addressed in later versions.

Transcript versions

1. Original live speech-to-text transcript — TXT: the original app export, with the uploaded contents preserved byte for byte and the original filename retained.
2. Version 1 — TXT: the first full edited draft, with 43 numbered original-wording comparison blocks (U01–U43) and uncertainty notes.
3. Version 2 — TXT: incorporates further corrections, including the name Zachary Goodale at U32. It retains 32 numbered comparison blocks, using their original U numbers so they can be matched to v1.
4. Version 3 — PDF: the Daniel-reviewed draft, including his foreword, further wording and speaker corrections, and explanatory context about documents discussed. Numbered comparison blocks have been removed; some uncertainty notes remain.
5. Final revised transcript — PDF: the final participant-reviewed version, with subsequent editorial corrections. Compared with v3, it changes “resigned from emergency” to “resigned from [our agency]” and changes the comma after “His family moved to Spanish Fork” to a period. The family-move sentence remains in the Austin Lund passage on the PDF page.

Publication and export corrections

On October 21, 2026, timestamp spacing in the edited text drafts was normalized, and the text links were configured to use UTF-8 so punctuation displays correctly. The previously published copy labeled original, which had normalized timestamp spacing, was replaced with the unchanged uploaded app export under its original filename. The earlier published copy was removed. These display corrections did not change the interview wording.

The v2 introduction has also been corrected to list the U32 name correction and report 32 remaining comparison blocks, rather than 33. Both PDFs had retained an internal document title referring to “version-2” during export; their metadata titles have been corrected to identify v3 and final respectively. These review-note and metadata corrections did not change the interview body or the PDF page layouts.