Roughly half the endnotes are post-1993 additions, and the endnote system doubled as the refuge for material Pietsch wanted cut.

This is material that was added after the early draft: absent from the 1993 manuscript, present in the published novel.

In the manuscript. The 1993 manuscript has footnotes at page bottoms, "most endnotes added later."

In the published book. This concerns the book as a whole rather than a single scene. (endnotes; endnote-scenes endnote-scene n024–endnote-scene n332)

Evidence. Independently corroborated. Documented by Steven Moore's first-draft collation (half added; footnotes-at-bottom); D.T. Max's biography of Wallace (favored evasions: condense, or move into endnotes, "where some of his favorite passages went to make their last stand").

Bears on. This difference bears on the following open questions: the macro-shape: ring, gasket, thematic chapters, dead-center design.

Further notes. See diff-endnote-system for the invention itself. The cut pressure on the notes is now primary-witnessed Michael Pietsch's December 1994 second-read letter: "I still hope you'll look for many, many notes you can cut. I've suggested xing more of them but will go into that another time." — the environment in which roughly half the endnotes were nonetheless added; the refuge held. Second-venue corroboration Farber's 2009 essay on editing Wallace: the book kept "more than 100 pages of (often footnoted) endnotes" while Pietsch persuaded Wallace to cut "hundreds more" — the added-under-cut-pressure shape, via the same Moore/Max record.