Skip to content
Welcome / Blog Archive / English / 2020-04-doc5 Do we know what a Rembrandt painting looks like?

2020-04-doc5 Do we know what a Rembrandt painting looks like?

It took some 150 years before, gradually, the 19th century experienced Rembrandt becoming the subject of more serious scholarship. Research began with the recording of anecdotes about his life, but increasingly a more factual approach on his work was adopted. As early as 1810 the first catalogue raisonné of Rembrandt’s etchings was published by William Esdaile. A first catalogue of paintings was established by John Smith, ‘dealer in pictures’, in 1836 (

As the century progressed, historical documents were retrieved in dribs and drabs by scholars, including the Amsterdam archivist Pieter Scheltema, whose sniffer-dog capabilities in the archives earned him the nickname ‘Piet Perkament’ (‘Parchment Pete’). Towards the end of the 19th century, knowledge of Rembrandt’s art was increased considerably, especially by Abraham Bredius and Cornelis Hofstede de Groot. In 1898 this resulted in the first exhibition dedicated to a single old master in the Netherlands.

In addition, other experts, such as Wilhem Bode and Wilhelm Valentiner, serve as founding fathers of the history of the art history about Rembrandt. Scallen (2004), 322, concludes that the significance of these four persons mentioned between around 1880 and 1930 lies ‘… in their recasting of connoisseurship as a professional activity, their shaping of connoisseurship as an activity promoted through public debate, and their development of modern modes of art historical communication.’ Their valuabe contribution, including a wealth of publications, brought life and volume to Rembrandt’s work. As a drawback in the development of art historian research, it has been submitted that these four men in fact monopolised the field of connaisseurship: pretending superior knowledge, supporting each others views and stifling views expressed outside of this inner circle, not unveiling the criteria for attribution, allowing a strong individual appreciation, entangled with emotional, national and financial interests (Levine, book review,; DOI: 10.3202/; Haitsma Mulier, book review in: Bijdragen en mededelingen betreffende de geschiedenis der Nederlanden, januari 2005, 120(4):658; DOI: 10.18352/bmgn-lchr.6309).

A stronger emphasis on such elements as cultural milieu and iconografy since the 1930s has been made by scholars such as Bauch, Schmidt-Degener, Benesch and Gerson. New methods of research via technology casted doubt about atributions made earlier. In a period of some seventy years (c. 1900 – c. 1970) nine catalogues raisonnés were published by art historians of good report. The confusing number of attributions called for an initiative to produce a final catalogue raisonné of Rembrandt paintings, by extensively using technical investigations (X-radiography, infrared and ultraviolet photography, chemical analysis of paint layers, dendrochronology, canvas thread count, etc.) to establish authenticity. This awareness and understanding resulted, in 1968, in the so-called Rembrandt Research Project (RRP). A projectteam of seven experts aimed to analyse documentation, techniques, and forensic research on Rembrandt paintings from his early years until his death.

Originally envisaged to take a decade, it ended up lasting forty-six years, ending in 2014, in a still unfinished result, however terminated with the retirement of Dutch art historian and connoisseur Ernst van de Wetering. At that time approximately one-quarter of Rembrandt’s oeuvre had not yet been investigated. The result of the project is published in six voluminous books, collectively known as ‘A Corpus of Rembrandt Paintings’. The first three volumes were published in 1982, 1986 and 1989 respectively. These caused a great deal of unrest among art historians, owners of work attributed to Rembrandt and curators of musea, especially since many works were no longer recognized as a Rembrandt. Because of the criticism, four of the original seven members of the investigation team had resigned. Van de Wetering, who was involved in the project from the start as an assistant, but later as a member of the project team, was placed in charge of the project in 1993.

The fourth and fifth parts of the RRP were given a different, non-chronological, design. When Van de Wetering withdrew from the daily management at the age of seventy-two in 2011, it was decided to close the project with the publication of the sixth part. As noted, at that time approximately one-quarter of Rembrandt’s oeuvre had not yet been investigated. In particular in this last part, which was published in October 2014, many of the conclusions from the first three parts were revised. The first five parts of the corpus can be viewed on the web pages of the Rembrandt Database (

A Corpus of Rembrandt Paintings in six parts, with various writers, has the following set-up:

Volume I, 1629-1631, 1982;

Volume II, 1631-1634, 1986;

Volume III, 1635-1642, 1989;

Volume IV, Ernst van de Wetering, Self-Portraits, 2005;

Volume V, Ernst van de Wetering, The Small-Scale History Paintings, 2011;

Volume VI, Ernst van de Wetering, Rembrandt’s Paintings Revisited, A Complete Survey, 2014;

The archives and documentation of the RRP have been taken over by the RKD – Netherlands Institute for Art History in The Hague. With funding from third parties a pilot initiative has been launched, called the Rembrandt Database, that will build on and supplement research from the RRP.

I understand that the project’s six-volume publications ‘A Corpus of Rembrandt Paintings’ is considered the definitive authority by all auction houses and dealers who work with works by Rembrandt and his studio. The sixth and final volume was presented by Ernst van de Wetering, the sole author of this Volume and final remaining member of the Rembrandt Research Project. In this volume (therefore as per 2014) Van de Wetering counts 340 autograph Rembrandts. Some sixty of them have now been accepted by Van de Wetering, not as newly discovered works, but as the results of reassessing these. Van de Wetering wished to purify Rembrandt’s oeuvre, as in his view earlier experts provided poor arguments.

This final volume, not surprisingly, initiated debate about the feasibility of the conclusive attributions and is severely criticised by for instance Michael Savage, see, by White (2015), as well as by Gary Schwartz, see

Back to my question: do we know what a Rembrandt painting looks like? It’s an art historian question, so I just point at an author writing four years ago and providing an overview of the organisation of the Rembrandt Research Project, its methodology and it results, see Krzyżagórska-Pisarek (2016), 23ff. Her conclusion (at p. 37) throws cold water: ‘Despite fifty years of research and all the technical investigations the confidence in Rembrandt connoisseurship is lower then ever. Today nobody seems to know what a Rembrandt painting should look like anymore’.