Who Painted It? How Museums Investigate an Artwork’s Identity

Category: Arts & Culture | Published: 10/14/2026

Museums identify an artwork’s maker by combining provenance, close looking, technical imaging, material analysis, and comparison—not by relying on one perfect test.

Museums investigate an artwork’s identity by combining several independent kinds of evidence: its ownership history, labels and documents, visible style and technique, physical construction, scientific images, material analysis, and comparison with securely documented works. No scanner flashes “genuine” or “fake.” A strong attribution is a reasoned conclusion that explains which clues agree, which conflict, and what remains uncertain.

That is why museum labels sometimes change. New records may appear, a painting may be cleaned, or technology may reveal an underdrawing or material that specialists could not previously see. A revised attribution is not necessarily a failure. It can be evidence that the museum is updating its conclusion when the evidence changes.

Authentication and attribution are related—but not identical

In everyday language, people often ask whether a painting is “authentic.” Museum specialists usually face a more precise question: who most likely made this object, where, and when? That process is called attribution.

The answer may be straightforward, but it can also be qualified. A catalogue might describe a work as “by” an artist, “attributed to” that artist, “workshop of,” “circle of,” or “follower of.” Institutions do not always use these labels in exactly the same way, so readers should check the museum’s catalogue notes. The wording communicates a level or kind of connection, not merely a price tag.

An artist’s signature helps only when it fits the rest of the evidence. Signatures can be added later, copied, altered, or used by a workshop. Even a genuine old canvas or wooden panel does not prove that the famous name attached to it is correct.

Start with the paper trail: provenance

Provenance is the documented history of an artwork’s ownership, custody, sale, exhibition, and movement. Researchers may examine auction catalogues, dealer stock books, collection inventories, shipping records, photographs, correspondence, exhibition labels, and marks on the back of a frame or panel.

A continuous, well-supported chain can connect an object to a particular place and time. A gap does not automatically mean forgery: private sales, lost archives, wars, inheritance, and changing record systems can interrupt the trail. Conversely, a polished certificate is not enough by itself. Researchers ask whether each document existed at the claimed time, whether names and dates match other records, and whether the document can be independently traced.

The Getty Provenance Index demonstrates the scale of this work by making archival records from sales catalogues and dealer stock books searchable. Provenance can also raise ethical and legal questions about theft, forced sales, colonial removal, or wartime displacement. Establishing who owned an object is therefore about more than naming its maker.

Close looking: what the object itself can reveal

Curators, conservators, and art historians first study what can be observed directly. They look at composition, brushwork, drawing habits, repeated motifs, spelling in inscriptions, changes made during painting, the type of canvas or panel, joins, nails, stretcher construction, damage, varnish, and previous repairs.

They then compare those observations with securely attributed works. Does the artist usually build faces in the same sequence? Are the hands, folds, highlights, or perspective handled similarly? Does the support match materials available in the proposed region and period? Comparison is valuable because artists develop habits—but it is not infallible. Artists experiment, assistants participate, copies imitate originals, and restoration can disguise the surface.

Specialists sometimes use the term connoisseurship for close visual judgment informed by deep familiarity with an artist, school, or period. Good connoisseurship forms testable hypotheses. It becomes stronger when documentary and technical evidence support it.

What can science see beneath the surface?

Conservation science helps investigators study layers and materials, often without removing a sample. The National Gallery of Art’s scientific research department describes a collaborative approach involving scientists, conservators, and curators. Its tools investigate artists’ methods and materials, support treatment decisions, and map pigments or binders with visible light, infrared radiation, and X-rays.

Common tools in a museum evidence file
Method	What it may reveal	Important limit
Magnification and raking light	Brushwork, cracks, surface texture, raised paint, repairs, and tool marks	Surface appearance may have changed through aging, cleaning, varnish, or restoration
Ultraviolet examination	Differences in varnish, retouching, and some materials through fluorescence	Interpretation depends on the material and cannot identify every alteration
Infrared reflectography	Some carbon-containing underdrawings, hidden changes, and features beneath certain paint layers	Not every pigment or drawing material responds clearly, and thick or unsuitable layers can block the view
X-radiography	Construction, dense pigments, nails, joins, changes, and earlier painted forms	The image combines layers into one projection and requires expert interpretation
X-ray fluorescence and related spectroscopy	Chemical elements that help specialists infer pigments and map their distribution	An element may occur in more than one material, so context and comparison remain necessary
Microscopy and tiny samples	Layer sequence, fibres, pigment particles, binders, coatings, and later additions	Sampling is limited and carefully justified because it removes a minute amount from the object
Tree-ring analysis	A date range for wood in a panel when its ring pattern matches a regional chronology	It dates the wood growth, not automatically the painting event or the artist

A useful example comes from the Getty’s study of Vincent van Gogh’s Irises. According to the Getty’s account of the examination, microscopy helped researchers inspect pigment mixtures; infrared reflectography and X-radiography revealed preparatory layers and changes; and macro X-ray fluorescence mapped chemical elements from which pigments could be inferred. The finding did not come from one image. Researchers compared several datasets with knowledge of Van Gogh’s other works.

Materials can challenge a proposed date

A pigment, binder, support, or manufacturing method may provide a chronological clue. If analysis finds a material that was not available until long after an artist died, the proposed attribution needs serious re-examination. Investigators still have to ask where the material occurs. A modern pigment in a small repaired area is different from the same pigment appearing throughout the original design.

The National Gallery in London documented a striking case in its “Fakes, Mistakes and Discoveries” research: a portrait once believed to be fifteenth-century contained pigments unavailable before the nineteenth century, while a later coating helped imitate age. The same project also showed how scientific analysis, conservation, connoisseurship, and historical research can correct older mistakes or identify collaboration within a workshop.

Material evidence is especially powerful when it rules out a claim, but compatibility is not proof. A forger can use old materials, and many artists working in the same place shared pigments, suppliers, panels, and workshop practices.

A museum-style investigation, step by step

Define the claim. Is the question about the artist, workshop, date, place, original design, ownership history, or whether parts were added later?
Record the object before intervention. Specialists document dimensions, condition, inscriptions, labels, front, back, frame, support, and visible repairs.
Build the provenance timeline. Each ownership or exhibition claim is linked to a traceable record, and gaps are marked honestly.
Compare securely documented works. Researchers examine technique, construction, materials, drawing habits, and historical context—not just subject matter.
Select appropriate technical methods. The question determines the tool. Infrared imaging cannot answer the same question as elemental mapping or tree-ring analysis.
Test competing explanations. Could an unusual feature come from an assistant, a copyist, later restoration, aging, or a change made by the artist?
Write a calibrated conclusion. The final catalogue entry explains the evidence, confidence, and unresolved questions. Future discoveries may change it.

Why one dramatic clue is never enough

Imagine a painting with a famous signature, old-looking cracks, and a nineteenth-century sale label. Those clues sound impressive, but each needs checking. Was the signature painted at the same time as the image? Are the cracks natural, or are they only in a later varnish? Is the label genuine, and can its catalogue number be matched to an archived sale?

Now imagine that infrared imaging reveals a careful underdrawing similar to documented workshop practice, the wooden panel fits the proposed period, and the pigments are historically plausible—but the ownership record begins two centuries later. The evidence supports further investigation, not certainty. A responsible conclusion might identify the most likely artist or workshop while clearly recording the provenance gap.

This is the same evidence-first habit used in other fields. In How Do Archaeologists Date the Past?, several methods strengthen a timeline when their limits are understood. In art research, agreement among documents, physical evidence, images, chemistry, and informed comparison is stronger than any single clue.

Try a safe classroom attribution case

Students can practise the reasoning without handling valuable objects or attempting chemical tests. An educator can provide public museum images and catalogue entries for three securely identified works plus one fictional “mystery” work assembled for the lesson.

Separate direct observations from interpretations.
Create a provenance timeline and flag every unsupported step.
Choose two technical methods and state exactly what each could test.
Compare at least three features with the reference works.
Write a conclusion using a confidence label such as strong, moderate, limited, or unresolved.
Name one new piece of evidence that could change the conclusion.

Do not clean, scrape, shine strong light on, or sample a real artwork. Conservation examination requires trained professionals and object-specific risk assessment. For research habits, How to Evaluate Online Sources shows how to check authorship, evidence, and corroboration before trusting a claim.

Reliable sources for further learning

The big idea

Museum authentication is not a contest between an expert eye and a laboratory machine. It is a collaboration among art history, archival research, conservation, chemistry, imaging, and careful reasoning. The best conclusion is not the most confident-sounding one. It is the conclusion that accounts for all the evidence, states its limits, and remains open to revision when a better clue appears.

← Back to Blog