Introduction I

Taste, connoisseurship and the history of the British Museum collection

Rembrandt’s drawings have been studied by modern scholars for more than a century, but in many cases their authenticity still arouses controversy. His paintings, though less numerous, have proved equally divisive. Only the corpus of his etchings has escaped a major attributional revision, not least because many are signed and dated, providing sufficient comparative material on which to base stylistic judgments.

The situation for the paintings and drawings remains unsatisfactory. To those involved in the debate, the reasons for the continuing controversy are all too obvious. The layman, however, views the situation with near incredulity. An explanation seems appropriate in the context of the present catalogue, which revises several long-held opinions – and some short-held ones – about the drawings attributed to Rembrandt in the British Museum.

Rembrandt’s paintings and drawings are poorly documented. The paintings, under the rules of the guilds in the 17th century, could be signed by him as they emerged from his workshop regardless of whether he had been personally involved in their execution. As for the drawings, in Rembrandt’s workshop they were rarely signed at all. Just three of the drawings in this catalogue, 108 of which were until the 1950s considered to be by Rembrandt, are clearly signed drawing(Cat. nos 7a, 31 and 38). [1]  The difficulties are exacerbated by the artist’s working practices. In particular, he was not in the regular habit of preparing his paintings with drawings. In this he differs from the great Italian and Flemish draughtsmen of the Renaissance and Baroque periods, such as Raphael, Michelangelo and Rubens, whose drawings can often be authenticated on the basis of their connections with finished, well-documented paintings. A nearer analogy is with Titian, whose surviving corpus of drawings also provokes animated debate, although it is small in comparison.

Taking these factors into account throws the problems that face cataloguers of Rembrandt’s work into relief. It might be – and indeed has been – argued that little consequence should be attached by serious researchers to questions of attribution and that an approximate location of an object within the correct school and period is sufficient to satisfy the demands of academic and even more general debate. But without an agreed canon of these works, there can be no reliable analysis of any aspect of Rembrandt’s true significance, or of his working methods, his ambitions as an artist, his predilections in subject matter or his influence as a teacher. Unfortunately, the interminable preoccupation with problems of authenticity has slowed or distorted research in these other areas. The same difficulties apply, of course, to the work of Rembrandt’s pupils, and it is to enable these fields of enquiry to expand that we believe the attempt to define the corpus of Rembrandt’s drawings and of artists of his school, to be necessary.

The attributional controversy is anything but new. In the case of Rembrandt’s paintings the problems can be traced to his own lifetime. [2]  Doubts about his drawings are less well documented, but a remark reported to have been made in 1710 by the collector N.A. Flinck (1646–1723), the son of Rembrandt’s pupil Govert Flinck and the owner of a sizeable group of Rembrandt’s drawings, that ‘drawings were difficult to know and that there was much deception in them’ suggests that he was aware of misattributions and perhaps of malpractices in the old master drawings market. [3]  Indeed, the presence of drawings by Rembrandt in early and distinguished collections – including those of Valerius Röver and Pierre Crozat – is no sure guide to their authenticity, [4]  and by 1796, when the British Museum had already acquired some drawings by Rembrandt, the artist’s oeuvre was so blatantly corrupted that the Rembrandt connoisseur Daniel Daulby could write: ‘if a picture possess any thing of the manner of Rembrandt, it is usually attributed to him, either to inhance [sic] the value, or to flatter the possessor’. [5]  If the paintings were attributed in such a cavalier fashion, then there is every reason to suppose that the drawings were treated in the same spirit. In short, works in a Rembrandtesque style, whether by Rembrandt’s pupils or later imitators, were fused with the master’s own creations.

Before the arrival of specialists and so-called ‘scientific’ connoisseurship in the mid-19th century, the degree of independent expertise available was limited. Traditional – or supposedly traditional – attributions were widely accepted and it was undoubtedly a simple matter to pass off paintings and drawings by Rembrandt’s pupils as by the master. Needless to say, such practices are difficult to document, but Daulby was sharp enough to notice that the pupils’ names rarely appeared in sale catalogues or other publications. [6]  Nor did the drawings’ owners, whether collectors or museums, have any incentive for enquiring into whether their Rembrandts should be ‘demoted’ and reattributed to his followers. This was a system without checks and balances.

It is therefore not surprising that, as a result of research in more recent times, sheaves of drawings formerly considered to be by Rembrandt now languish, little regarded, in most of the old-established European collections, including the British Museum. In Munich (at the Staatliche Graphische Sammlung), where the collections can be traced to the first half of the 18th century, a small group of authentic sheets is joined by numerous drawings by pupils and followers. These were frequently ‘enhanced’ by a crude, later hand, which also added false Rembrandt signatures. [7]


  • ^ [1] - One hundred eight British Museum drawings were catalogued as by (or retouched by) Rembrandt in the last catalogue raisonné of his drawings compiled by Otto Benesch (Benesch, 1973). Cat. no. 11 has a fragment of a signature, while the signature on Cat. no. 10 has been doubted, although in my view it is probably authentic, but rubbed.
  • ^ [2] - The inability of early collectors to distinguish between the work of Rembrandt and Lievens is well documented – see Corpus, I, 1982, under nos A12 (Simeon in the Temple, Hamburg, probably the picture called ‘Rembrandt or Lievens’ in the Stadholder’s inventory of 1632; and nos A38–9, the Berlin Minerva and Rape of Proserpina, both attributed to Lievens in the same inventory. See E. van de Wetering in Corpus, II, 1986, 60ff., for other early problems of authenticity.
  • ^ [3] - See Munich, 1973, 151–2: Z.C. von Uffenbach wrote that N.A. Flinck gestünde zwar, dass die Handrisse schwer zu kennen, und viel betrug dabey sey.
  • ^ [4] - See Schatborn, 1981 and Broos, 1985. The former believed that the inscriptions on drawings from the Ten Cate collection might be reliable, but Schatborn and I together discovered a study by Philips Koninck in the Albertina with this provenance, but which had been inscribed with Rembrandt’s name (Sumowski 1373x).
  • ^ [5] - Daulby, 1796, xiv.
  • ^ [6] - Daulby, ibid.
  • ^ [7] - See Wegner’s catalogue (Munich, 1973) and the articles by Burmeester and Renger, 1986 and 1985–6 and in Vignau-Wilberg (ed.), 2003, 45–62 (‘Händescheidung im Nahen Infrarot: die Münchner Rembrandt-Apokryphen’), as well as Exh. Munich, 2001–2.