In Mainz, between 1454 and 1455, the printer Johannes Gutenberg and his financial backer Johann Fust bring off the press about 180 copies of a two-volume Latin Bible, set with movable type: the first work of this scale printed mechanically in the West.
The text it carries, the Vulgate, is available today free of charge, in full, in identical form, in any library and on any screen. And yet one of those volumes, or even a single detached leaf, changes hands for millions of dollars: so it is not the text that is being paid for.
the estimated value of a complete copy; no sale of this kind since 1987
What is being paid for is not the reading of the text: it is a precise object's place in a census. Since the 19th century, bibliographers have kept an inventory of every known copy of this Bible, whether a complete volume, an incomplete volume, or a single leaf, giving it its own identity and as full an ownership history as can be traced.
Being entered in that census, with a verifiable record, is what turns an old book into proof.
When a leaf or a volume resurfaces, found in an attic, in a binding that was serving as mere board, or in the holdings of a religious institution, it is checked against the copies already on record: measurements, paper watermarks, rubrication marks, missing pages, binding. If it matches a known copy that had gone missing, it is attached to that copy's record; if not, a new one is opened for it.
Insurers, museums arranging a loan, and auction houses all rely on that record, and on it alone, to set a value.
The original print run falls, depending on the source, between 160 and 185 copies. In 1985, the Hungarian bibliographer Ilona Hubay records 47 of them, with their owners: two more will later be found in Russia, bringing the figure usually cited today to 49.
Of those 49, 20 to 25 are judged complete, depending on the criterion applied: even a rigorous census rests on a convention.
How to read it: each bar counts the same Bibles at a different moment in their history, without a strictly proportional scale. Indicative diagram. The loss between the first and second bar unfolded over five centuries; the one between the second and third mostly comes down to the definition of “complete” that is used.
The trajectory starts badly for its own inventor. On 6 November 1455, Fust sues Gutenberg for repayment of a debt of 2,026 guilders, interest included, on an original loan of 1,550 guilders made in 1450. The judgment goes against Gutenberg, who most likely loses to Fust the workshop and equipment with which the Bible had just been printed.
From that initial rupture onward, each copy follows its own documented path: private chapels, dispersed religious holdings, collectors, then institutions — the Morgan Library alone now holds three, more than any other library in the world.
In 1920, the dealer Joseph Sabin buys at auction a copy missing 50 leaves out of 643: unsellable as a complete volume.
The bookseller Gabriel Wells buys it and makes a decision that scandalises part of the trade: breaking it apart leaf by leaf rather than seeking a single buyer.
In 1921, Wells sells close to 600 separate leaves, and gives the last one, unsold, to the New York Public Library.
What makes this story coherent is that it rests on no disputed attribution and no legend reconstructed after the fact: every link carries a name, a date, and a verifiable record — the Mainz workshop, Fust's loan, the 1455 judgment, the bibliographic census, Wells's break-up of the book.
Unlike a painting whose value rests on a school's attribution, the value of this Bible rests on a file that can be consulted, piece by piece, without ever having to settle an expert dispute.
This work is dated with a precision rare for an object of the 15th century: a handwritten note added by a rubricator on a known copy gives 15 August 1456 as the date its hand-decoration was finished, and the study of the paper and the wear on the type lets specialists place the printing of each copy within a few months.
A document dated by a third party, and the physical examination of the object itself: two proofs that, since 1455, have never once been caught out.
The real value rests on the census record: the full set of surviving witnesses, known, documented, and shrinking only very slowly. The price now forms only on what still changes hands: separate leaves.
That market exists precisely because Gabriel Wells chose, in 1921, to manufacture the very scarcity he was selling: the census can be entirely right about overall rarity, while the price measures the convenience of owning a fragment rather than the survival of the whole object.