Anatomy of a single revision.

Mindfield Guides runs a periodic ingestion of the Psychedelic History Archive — a curated open archive of primary documents and researcher biographies in the psychedelic literature. The May 2026 sweep surfaced a researcher card for someone whose name was not in any of the Ayahuasca volume's existing seventeen chapters: Luisa Agusta Rebeca Gambier de Álvarez de Toledo (1915–1990), described as "an under-recognized pioneer of psychedelic therapy in 1950s Argentina."

The user flagged the card and asked the working question that every research session should start with: does this actually add anything, or is it already covered?

Before doing any new research, the existing seventeen chapter drafts have to be audited. There is no point investigating a "gap" that turns out to be already covered, and there is a real risk of redundancy or contradiction if the volume already has something to say on the same topic.

The Ayahuasca volume already cites several Argentine sources heavily — Berlanda and Viegas, the Buenos Aires psychoanalytic accounts in Chapters 5, 7, and 8. So it would have been entirely plausible for this lineage to already appear under a different name.

Now the real work. The archive card pointed at a 2023 article in Chacruna Institute by the Sorbonne historian Zoë Dubus titled "LSD and Ayahuasca in Argentina: The Pioneering Work of a Psychoanalyst in the 1950s." That article was the most accessible English-language synthesis, but it is a secondary source. The question is what the primary record actually supports.

Reading Dubus carefully turned up a critical distinction that the archive's short description had flattened. Álvarez de Toledo's clinical psychedelic-therapy work was with LSD and mescaline, not ayahuasca. Her ayahuasca work was a single 1960 paper that Dubus describes as "more anthropological." Two different kinds of engagement, with two different evidentiary weights.

At this point a working draft of how the section might read was held against the evidence. The first draft framed her as "the first Western clinician to use ayahuasca therapeutically." That framing collapses on contact with the primary record. The narrative that survives the collapse is more specific, smaller, and more accurate.

Before the paragraph could be written, the underlying 1960 paper had to be confirmed real — not in the form Dubus had summarized, but in the bibliographic form a reader could verify. That meant checking PubMed for indexing, checking Encyclopedia.com's biographical record for the complete citation, locating the document on PEP-Web (the standard archive for psychoanalytic literature), and cross-checking against Scholten's 2012 scholarly study of the surrounding LSD lineage.

Only after the underlying record was verified did the actual prose get written. The new section — ~285 words under the heading "Buenos Aires, 1959 to 1961" — was inserted into Chapter 2 between the existing section on the Beat-era literary discovery of ayahuasca and the existing section on the 1993 Hoasca Project in Manaus. The placement does narrative work: it explains in part why the Western record on ayahuasca has a long quiet stretch between mid-century and the 1990s.

Adding a paragraph is the easy part. The hard part is making sure that every (Author Year) marker in the new prose actually points to an entry in the chapter's citations.md file — and that no existing reference broke when the file was edited. We do this with a small audit script, not by reading the file linearly.

The first integration left three honest gaps in what was known. Rather than paper over them, they were written down as open questions and re-investigated in a second pass. This is a discipline that matters — the gaps go in a deep-research note inside the chapter folder so the next reviewer (human or otherwise) can see exactly what was unverified.

The visible result of this entire session is a single new ~285-word section in Chapter 2 of the Ayahuasca volume, four new bibliographic entries in its references, and a corrected inventory header. The invisible result is roughly two dozen tool calls, four parallel-source database checks, a falsified initial framing, a programmatic citation audit, and a documented deep-research note. Most of it never appears in the book.

The Veigl card from the same archive ingestion was also reviewed in this session and dropped — the volume already cited the 1750s Jesuit material in Chapter 1, and the unused passages were not about the brew. Dropping a candidate is part of the work.