LeCun et al.'s two 1989 ZIP-code papers appeared in a bimodal order: the NIPS conference paper was presented first (Nov 1989), the Neural Computation article was printed first (Dec 1989), and the NIPS proceedings volume only appeared in 1990

lecunbell-labsbackpropagationnipsneural-computationzip-codeshistory-of-mlbibliography

LeCun's Bell Labs group published closely related handwritten-digit work twice in 1989: the NIPS conference paper "Handwritten Digit Recognition with a Back-Propagation Network" (Advances in NIPS 2, pp. 396–404) and the Neural Computation journal article "Backpropagation Applied to Handwritten Zip Code Recognition" (claim-lecun-1989-first-practical-recognition). "Which appeared first" has two correct answers, and citing "LeCun et al. 1989" without specifying the venue gets the sequencing backwards:

• By presentation: the NIPS conference paper came first — the conference was held Nov 27–30, 1989 in Denver (dblp: "NIPS Conference, Denver, Colorado, USA, November 27-30, 1989. Morgan Kaufmann 1990").
• By printed publication date: the Neural Computation article came first — in print December 1989 (Crossref published-print 1989-12, DOI 10.1162/neco.1989.1.4.541; OpenAlex normalizes this to 1989-12-01), Volume 1, Issue 4, pp. 541–551.
• The NIPS 2 proceedings volume that formally contains the conference paper was not printed until 1990 (dblp "Morgan Kaufmann 1990"; Open Library first_publish_year 1990) — a year after the conference, even though the NeurIPS archive retroactively labels it "NIPS 1989."

So: NIPS presentation (Nov 1989) → Neural Computation print (Dec 1989) → NIPS proceedings volume in print (1990). The "1989" on the conference paper is the conference year used as a citation label, not the volume's print date. See moc-backpropagation-origins.