Scientific investigation · 1968 -2014
X-rays, dendrochronology, study of paintings and the connoisseur's gaze: half a century of research has transformed our idea of Rembrandt - without closing all the debates.
Immediate response
The project didn't just remove tables: it changed the question
The Rembrandt Research Project (RRP) was established in the Netherlands in 1968 to systematically examine paintings attributed to Rembrandt. His initial ambition was to distinguish originals from copies, studio works and imitations. But the investigation quickly revealed a more complex reality: Rembrandt ran an active workshop, his students shared materials and practices, and some paintings may have been started, corrected, or completed by several hands.
A timeline in five turns
Resume the work table by table
Josua Bruyn, Bob Haak, Simon Levie, Pieter van Thiel and Ernst van de Wetering were among the first members. The project was born in a context where the catalogs of Wilhelm von Bode, Cornelis Hofstede de Groot and Abraham Bredius had gradually expanded the painted work. The objective is to visit the collections, gather the files and apply consistent criteria.
System A, B and C
The paintings are classified between accepted works, uncertain attributions and rejections. This method makes the reasoning readable, but also gives the impression of definitive verdicts. The first three volumes cover 1625 -1642 and carefully document provenance, x-rays, support, material condition and style.
Of the commission with regard to Van de Wetering
Internal differences and criticism over certain downgrades lead to reorganization. Ernst van de Wetering, former technical assistant of the project, takes a central place. It places greater emphasis on the creative process, workshop practice and the possibility of re-examining the RRP's own conclusions.
Self-portraits and small history paintings
Volume IV deals with self-portraits; volume V small history paintings. The structure gradually abandons the old rigid classification. Previously rejected paintings can be readmitted, proof that the Corpus is a moving research rather than a fixed list.
A revised summary
Rembrandt's Paintings Revisited: A Complete Survey brings together 350 paintings considered by Van de Wetering as autographs or executed by Rembrandt with a student. The institutional project is coming to an end, while the technical documentation continues to live in The Rembrandt Database and museum laboratories.
How is a painting attributed to Rembrandt?
The support
Dendrochronology dates wood and can connect panels cut from the same tree. The study of the threads of a canvas sometimes brings together several paintings from the same scroll. This places a work in a material environment, without automatically identifying the hand that painted it.
The invisible layers
X-ray, infrared reflectography, X-ray fluorescence and samples reveal preparation, compositional changes and pigments. Repentance can signal an invention in progress; a mechanical copy reproduces more often a solution already decided. But a student could also modify its composition.
The way of painting
Researchers observe light construction, key order, impasto, glaze, and variable precision. A signature is never sufficient proof: it can be added, reinforced, or imitated.
The context
Origin, old inventories, formats, costumes and reports with drawings or engravings complete the investigation. The strongest attribution is where material data, history and stylistic judgment converge.
Starting from solidly documented works
To recognize a hand, a comparison core must be established. Works dated, documented or linked to a known order provide benchmarks for the underlying design, pigments and workmanship. THERE Anatomy lesson already shows a very personal orchestration: each face exists, but the look, the diagonal of the corpse and the demonstration of Tulp form a single action.
The danger, however, would be to establish a single style as standard. Rembrandt changes depending on the scale, date, order price and function of the painting. A smooth touch in a 1632 portrait does not invalidate the free impasto of a late work.
A painting is not a fingerprint
Science can rule out an impossible date or reveal an unusual construction; it does not produce a "Rembrandt test". Common materials circulated in the workshop and students learned precisely to imitate the teacher. Conversely, an anomaly can correspond to an authentic experiment.
The restoration of the Night Watch also reminds us that we look at objects transformed by varnishes, cleaning, cutting and aging. To attribute is first of all to distinguish what belongs to the creation of the 17th century from what the following centuries added or altered.
Three files that show why verdicts evolve
Saul and David : dismantled, assembled, rehabilitated
Long admired as a Rembrandt, the painting was discarded by the RRP in 1969. Restoration and research conducted by the Mauritshuis later established that the painting had been cut into fragments, reworked and completed before being reassembled. This story explained some of the inconsistencies once considered suspicious.
In 2015, the museum presented the work as a Rembrandt again, painted in two phases. The episode is exemplary: a confusing material state can mask the coherence of the original process. Reattribution does not remove all dating issues, but it relies on the convergence of technique, style and physical reconstruction.
The Polish Rider : masterpiece and zone of uncertainty
This mysterious rider is neither an identified portrait nor a clearly named historical scene. Van de Wetering defended the attribution to Rembrandt, while other scholars considered student participation, notably Willem Drost. The Frick Collection retains the attribution to Rembrandt.
The record shows that disagreement does not mean "science knows nothing". Support, pigments and x-rays may be compatible with the workshop; the delicate point then remains the part of judgment made on quality, consistency and handling. The Frick Collection today attributes the painting to Rembrandt; competing hypotheses are part of the story of its reception.
The false choice between "master" and "false"
A 17th century studio work is not a modern fake. It may have been designed before Rembrandt's eyes, with its models, accessories and materials. Some students produced variations intended for the market; the master could correct a head, reinforce a light or let a painting go without intervening.
Vocabulary matters: "Rembrandt" indicates a work held for autograph; "Rembrandt's workshop" an execution in its immediate environment; "circle" a proximity without proof of direct learning; "follower" an often more distant stylistic dependence; "after" a revival of a known composition.
The RRP's main legacy is not a perfectly drawn boundary, but a public way of exposing evidence, doubts and changes of mind.
What is established - and what is still interpretation
| Question | Established or strongly documented facts | Interpretation to be handled with caution |
|---|---|---|
| Origin of the project | The RRP was founded in 1968; six volumes of Corpus published from 1982 to 2014. | The idea that the project would have provided a "final" list: its authors themselves revised conclusions. |
| Number of works | Volume VI retains 350 paintings as works by Rembrandt or Rembrandt with a student. | Present 350 as a universal total: catalogs and museums do not always adopt exactly the same positions. |
| Science | Analyzes can date a medium, characterize a preparation and visualize changes. | Claim that a machine directly identifies the author. Human interpretation remains essential. |
| Atelier | Rembrandt trained students and directed a production in which models, supports and compositions circulated. | Assign any difference in quality to a student: the state of conservation and the teacher's objectives vary. |
| Signature | A signature must be studied materially and historically. | Take it as an automatic certificate of authenticity. |
Compare periods without turning style into a verdict
To compare periods without reducing Rembrandt in a single way, compare objects of comparable date, scale, and function. A difference in finish must be examined with the material history of the painting.
| Question of comparison | What she avoids |
|---|---|
| Is this a small study, a trony, or a commissioned portrait? | Require the same degree of finish everywhere. |
| Does the chosen reference belong to a nearby period? | Treating an evolution in a manner as evidence of falsity. |
| Are the passages compared in a readable state? | Attribute to the painter what may be wear, varnish or rework. |
Start by describing the visible discrepancies, then look for the technical files and current formulation of the museum. There National Gallery of Art exposes attribution difficulties in Rembrandt's studio. This reading complements the case studies, without transforming the grid into an authenticity test.
Reproductive references for further study
Read a "Rembrandt" yourself in six steps
Read the cartel to the word
Locate the qualifiers: assigned to, workshop, circle, follower, according to. Consult the date the notice was updated online.
Look at the support
Panel or canvas? Original or fragment format? A seam, an extension or a cut can disrupt the composition.
Follow the light
In Rembrandt, she constructs action: she does not just illuminate objects. Ask yourself what it connects and what it deliberately leaves undecided.
Approach the surface
Observe the alternation between impasto and thin layers, incisive details and open areas. A photographic reproduction often crushes this topography.
Look for changes
A moved hand or a recaptured silhouette can testify to a process of invention. Museum technical records show what the naked eye cannot see.
Compare without simplifying
Choose a work close in date, function and dimensions. Comparing a small youth panel to a large late canvas easily leads to false verdicts.
Exercise: material, costume and identity
Start with the face: the wrinkles are not drawn one by one, they emerge from an irregular light. Then move on to metal, feathers and fabrics. The differences in finish do not necessarily reflect several authors; they prioritize attention.
Finally, don't look too quickly for a historical figure. Many costumed heads are tronies, studies of character and expression rather than named portraits. This is precisely the type of object where the old market, student papers and workshop variations complicate attribution.
Where to see the works and consult the research?
Amsterdam
The Rijksmuseum notably preserves the Night Watch, self-portraits and numerous documents. The Museum Rembrandthuis recreates the setting of the workshop and offers a scientific portal.
The Hague
The Mauritshuis presents the Doctor Tulp's anatomy lesson and Saul and David, whose restoration file perfectly illustrates a reallocation.
New York
The Frick Collection preserves The Polish Rider. The Metropolitan Museum publishes technical notices and research on its Rembrandt collection.
Online
The Rembrandt Database provides access to the first five volumes of the Corpus, technical files and a bibliography. Volume VI remains the final summary of 2014.
The work before the label
Explore Rembrandt with clearly marked attribution
Our collection distinguishes autograph works from paintings by studio, followers or after Rembrandt. For this topic, we provide a direct link to the general collection and, as a relevant study, to Dr. Tulp's Anatomy Lesson, without presenting the old attributions as originals.
Related articles
Frequently asked questions
What is the Rembrandt Research Project?
A Dutch research team founded in 1968 to systematically study paintings attributed to Rembrandt, combining stylistic knowledge, archives and scientific examinations.
Did the RRP remove half of the Rembrandts?
The formula is too simplistic. The project removed many works, but it also readmitted some. Totals vary depending on the catalog taken as a starting point and the definition of works partially executed in the workshop.
How many of Rembrandt's paintings are recognized today?
Volume VI of Corpus, published in 2014, selects 350 as works by Rembrandt or Rembrandt with a student. This figure is a major scholarly position, not an obligatory unanimity of all museums.
Does a signature prove authenticity?
No. A signature can be original, late, reinforced or added. It must be compared to the pictorial layers, the provenance and the entire technique.
What does "Rembrandt’s workshop" mean?
That the painting is considered to be produced in his studio by a student or collaborator, generally during his lifetime. This means neither modern forgery nor entirely autograph work.
Can science identify Rembrandt's hand?
It can date, compare and sometimes exclude. It reveals materials and processes, but the identity of the hand results from a synthesis with art history and visual judgment.
Why are the attributions still changing?
Because the works are restored, new scientific images appear, archives are discovered and the workshop's operating models evolve.
Saul and David is it really from Rembrandt?
The Mauritshuis attributes it to Rembrandt after extensive research and restoration presented in 2015. The painting was painted in two phases and had undergone significant material alterations.
The Polish rider is it disputed?
Its attribution was discussed: specialists proposed participation or attribution to a student. The Frick Collection presents it today as a work by Rembrandt.
Main sources
- The Rembrandt Database - A Corpus of Rembrandt Paintings, volumes I to V.
- The Rembrandt Database - catalogs of the work and volume VI (350 paintings).
- Museum Rembrandthuis - scientific portal and bibliography.
- Museum Rembrandthuis - Ernst van de Wetering and the RRP.
- Mauritshuis - Saul and David, collection notice.
- Rijksmuseum - The Night Watch.
- The Frick Collection - The Polish Rider.
- Ernst van de Wetering, A Corpus of Rembrandt Paintings VI: Rembrandt's Paintings Revisited - A Complete Survey, Springer, 2014.
The responsibilities indicated follow the notices of the institutions cited and are formulated with their reservations. Dates and dimensions may change after restoration or documentary re-examination.
0 comments