# Two Objects Here Can Prove Who Made Them
> Of 257 published records, 156 carry an accepted attribution and two carry one the object itself documents. Exactly one carries an attribution the holding museum has withdrawn. The distance between those three numbers is the subject.
> The Study - 2026-09-09
> Canonical HTML: https://iconotheca.com/study/two-objects-here-can-prove-who-made-them/
## Key points
- Across the 257 published records, this archive grades attribution seven ways: 156 accepted, 43 disputed, 39 unknown, 8 workshop, 8 attributed, 2 documented and 1 rejected.
- Only two records carry an attribution the object documents itself. The Ardabil Carpet knots the name Maqsud of Kashan and the year 946 into its own field, and the Archimedes Palimpsest carries a colophon giving the scribe as Ioannes Myronas.
- Neither of those two is entirely clean. The palimpsest colophon is in a later hand than the manuscript it closes, and published sources give the date as 13 April and as 14 April 1229.
- Exactly one attribution in the archive has been withdrawn by the institution holding the object. Wilhelm von Bode bought The Man with the Golden Helmet as a Rembrandt in 1897, and the Gemaeldegalerie withdrew the attribution in 1986 after technical and art-historical examination.
- Withdrawal is rare in the record because a demoted object usually leaves the record, not because attributions are rarely wrong. The Berlin canvas kept its wall because the museum said plainly that it is not a fake.
- The question does not stay closed. On 2 September 2026 an AI analysis was reported as putting Rembrandt's hand in the Berlin picture at 96 per cent probability, and the museum's 1986 position is unchanged.
## An attribution is a claim about a person, and almost none of them come from the object
## The carpet signs itself, and gives the year
## The prayer book names its scribe, and the day is still argued over
## One hundred and fifty-six accepted attributions rest on agreement
## Exactly one attribution here has been withdrawn
## Attributed is a weaker claim than accepted, and the catalogue says so
## The most expensive picture in this archive sits in the disputed column
## The withdrawn attribution did not stay withdrawn quietly
## How to read an attribution line
## Why this archive is free to read
Two objects in this archive can prove who made them, and both do it in their own material: a name knotted into a carpet and a colophon under an erasure. The other 255 attributions are somebody's judgement, however old, however distinguished and however widely repeated, and the difference is the first thing a reader should establish and the last thing a catalogue entry makes obvious.
## Questions
**What does a documented attribution mean in this archive?**
It means a contemporary record on or with the object names its maker, rather than a scholar proposing a name from style. Only two of 257 published records qualify: the Ardabil Carpet, which knots the name Maqsud of Kashan and the year 946 into its own field, and the Archimedes Palimpsest, whose colophon names the scribe Ioannes Myronas.
**How many attributions here are disputed?**
43 of 257 published records carry a disputed attribution, meaning named authorities disagree in print. A further 39 are graded unknown, where no maker is named and none can be, and 8 are graded attributed, where a name has been proposed on style.
**Why is only one attribution marked rejected?**
Because an object that loses a great name usually stops being lent, photographed and published, so it drops out of archives built on published evidence. The Man with the Golden Helmet is here because the Gemaeldegalerie kept it on display and kept explaining the 1986 withdrawal.
**Who says The Man with the Golden Helmet is not by Rembrandt?**
The Gemaeldegalerie in Berlin, which holds it. Its own catalogue record states that in 1986, after investigation by technological and art-historical methods, the attribution was withdrawn once and for all and the painting was designated a work by Rembrandt's circle.
**Did an AI reattribute the Berlin Rembrandt in 2026?**
An AI analysis was reported on 2 September 2026 as putting Rembrandt's hand at 96 per cent probability, and the report was followed up in the German press on 4 September. The museum's catalogue position, based on examination of the physical painting, is unchanged.
## Sources
- [1] Iconotheca, records.json, the archive's own published corpus. The attribution status distribution across the 257 records carrying status published was measured directly on 9 September 2026: 156 accepted, 43 disputed, 39 unknown, 8 workshop, 8 attributed, 2 documented, 1 rejected.
- [2] Victoria and Albert Museum, collection record for The Ardabil Carpet, museum number 272-1893. Read for the inscription and the museum's own translation, the identification of Maqsud Kashani as the person in charge of production, the year 946, the dimensions, the knot density of some 5,300 knots per ten square centimetres, and the 1893 purchase from Vincent J. Robinson and Co. for GBP 1,750.
- [3] The Archimedes Palimpsest Project, 'The History of the Archimedes Manuscript'. Read for John Lowden's 2002 ultraviolet reading of the colophon on folio 1 verso and the date of 13 April 1229, and for the dating of the Archimedes undertext to the second half of the tenth century.
- [4] William Noel, 'Archimedes and Company', British Academy. Read for the colophon being in a later hand, its wording being consistent with a colophon without being explicit, the name Ioannes Myronas, and the date given there as 14 April 1229, which is the one-day divergence recorded in the text.
- [5] Staatliche Museen zu Berlin, museum-digital object record for Der Mann mit dem Goldhelm, Gemaeldegalerie, inventory 811A. Read for the circle-of-Rembrandt attribution, the 1650 to 1660 dating, canvas, 67.5 by 50.7 cm, Wilhelm von Bode's 1897 acquisition as an autograph Rembrandt, and the museum's statement that the attribution was withdrawn once and for all in 1986 after investigation by technological and art-historical methods.
- [6] Christoph Droesser, 'KI haelt den Mann mit dem Goldhelm fuer einen Rembrandt', 2 September 2026, the author's own posting of his Deutschlandfunk item. Read directly for the sentence reporting that an AI sees Rembrandt's hand with 96 per cent probability in a picture classed since 1986 as from Rembrandt's circle.
- [7] Ingeborg Ruthe, 'KI haelt den Goldhelm fuer einen Rembrandt - Berlin widerspricht', Berliner Zeitung, 4 September 2026. Headline, standfirst, byline, publication date and the opening reference to the Deutschlandfunk report of 2 September were read; the remainder of the article is behind a paywall and was not read.
- [8] The Metropolitan Museum of Art, collection record for the Helmet all'Antica, object 21974, accession 04.3.202, read through the Museum's public collection API on 9 September 2026. The artist prefix field returns the literal string 'Attributed to' against the artist name Filippo Negroli, with the date about 1532 to 1535, the medium steel, the culture Italian, Milan, and the credit line Rogers Fund, 1904.