Lyric Witnesses: Printed, Sung, Remembered, Reused

Which lyric witness?

A serious archive cannot ask only “what are the lyrics?” It must ask which lyric witness? The Owner’s Manual, a live vocal, an earlier demo, a prose publication and a later reused line may all disagree while remaining historically valid.

Printed lyric versus sung lyric

LTAP production documentation already makes this distinction visible: vocal overlaps, discernibility, ad-libs and timing are discussed separately from the printed lyric fields. Tentacoli is a clean example. The voice can create a version of the text that the page does not fully represent.

Memory is an editor

the Solution to Art: a Song is a Kind of Engine explicitly presents remembered song material and admits that what follows is merely what could be remembered. The broader archive contains lost lyrics later found, damaged recall, lines reused after long gaps and phrases whose corrupted versions become useful. Forgetting can destroy evidence; it can also create a new witness.

Lost and found

Forbidden Lyrics: Significant Archival Discovery is almost comically small, but conceptually exact: lyrics to a “forbidden audio” are announced as lost and then found. That trivial archival event exposes a major property of the system — preservation materially affects which future objects can be made from the past.

OCR is not the author

The present EGOTERRORISM quantitative pass uses OCR-derived liner recovery for 51 JQD-authored track slots. Those figures are useful at project scale but are not treated as a diplomatic edition of every line. The visual liner remains the higher authority when exact wording matters.

A proposed witness hierarchy

1. direct official text or source document;
2. contemporaneous publication witness;
3. recorded performance where audible;
4. later authorial reconstruction;
5. OCR / machine recovery;
6. editorial inference, clearly labeled as such.