The Donor's Deadline: Attributing an Unsigned Dutch Panel

Ranking: first the underdrawing, because it is closest to the artist's hand and can be compared with securely attributed drawings; second the maker's mark, but only after checking guild registers and whether it was added later; third dendrochronology, because it gives a terminus post quem, not an author; fourth the smalt cobalt content, because it dates practice broadly and many workshops used similar pigments. The ranking is provisional. A felling date of 1618 only proves the panel could not have been painted earlier; it cannot identify the painter. The underdrawing becomes stronger if it matches a secure work in the same hand, but weaker if the workshop used pattern drawings. The maker's mark could overturn the ranking if it is traced to a documented panel maker supplying a specific workshop. To shift my conclusion I would want a secure signature under ultraviolet light, a documented provenance link, comparative infrared images from at least three accepted works, and an independent second connoisseur's opinion. No single result proves authorship; convergence matters, and the absence of a signature is not proof of anything.

The wording should separate observation from inference. Example: 'Attributed to [name], Portrait of a Woman with a Lace Collar, c. 1625–35. Oil on oak panel. The panel's felling date, no earlier than 1618, and its smalt pigment are consistent with this date. Infrared reflectography reveals an underdrawing that differs from the workshop's known practice; the attribution therefore remains provisional and rests on [specific comparison].' Avoid 'by' unless the evidence is strong; use 'attributed to', 'workshop of', or 'circle of' as the evidence warrants. The public label should give viewers enough to understand uncertainty without becoming a technical report, while the fuller reasoning goes in the catalogue entry and internal file. Do not mention the donor condition in the public text, because that would make a scholarly judgement look like a fundraising transaction. If the evidence is genuinely split, a date range and an honest 'attribution uncertain' is more defensible than a false certainty that will be tested by future scholars, lenders, and insurers.

The director should not allow a donation to determine an attribution. A conditional gift that requires a named master creates a conflict of interest and risks misleading the public. The correct process is to thank the donor, explain that attribution is an independent scholarly judgement, and offer to proceed with the gift unconditionally or to accept funding for research without an attribution condition. If the donor withdraws, the decision should be documented. The director should also seek external review, because the senior curator's insurance point is real: reattribution changes loan fees, security, and expectations. A defensible response is: 'We cannot guarantee an attribution; we can guarantee a rigorous process and a public record of our reasoning.' That protects the gallery's integrity and the donor's philanthropic intent. If the family wants certainty, the gallery can offer a public seminar on attribution, but it should not sign a deed that turns a research question into a contractual promise.