3.11 Weave matches
The study of the canvas support of paintings can help to determine sometimes unexpected relationships between paintings. A weave match suggests that two canvases were cut from the same bolt, and thus demonstrate that the two must have been in close physical proximity. This can help to confirm an attribution of a painting to a specific artist and his workshop, or to date it in a specific period, or to establish direct connections in the form of pendant pairs or part of the same series or the same commission.
One of the conclusions of Van de Wetering in 1986 on the canvases of Rembrandt paintings is that only a small number of canvases tend to come from the same bolt. Most such groups consist of two paintings, one of three paintings and one of five paintings. The paintings within such a group seem to have been painted within two years. This indicates that the painter bought his canvases in small batches and used them over a short period of time.
Research on Vermeer’s canvases shows a similar behavior. As we will see in chapter 6, of all Vermeer's paintings, there are three groups of two paintings and one of four paintings where canvases originate from the same bolt. Vermeer provided only a few paintings with a date, which makes the determination of an exact date of creation of each painting difficult. Overall, the weave matches seem to involve paintings which were made close to each other in time. There is one exception within the group of four. While three paintings in this group seem to have been made between 1662 and 1664, one is usually dated several years later, in circa 1670-1671. There is, as yet, no clear explanation for this time difference. The examination of canvases by other painters may help to find an explanation. It would be especially interesting to investigate the canvases of other Delft painters to see if there are canvases of these painters, which come from the same bolt as used in the paintings of Vermeer. This research should also include local primers, a nearly unexplored profession.
The results of the investigations of canvases can confirm assumptions on the attribution to a specific painter and his workshop. Van de Wetering’s study has demonstrated that in several cases a painting, which was not painted by Rembrandt himself, has in all probability been made in Rembrandt’s workshop, because the canvas originated from the same bolt as the canvases of secure paintings by Rembrandt. A good example of such a workshop painting is a free copy after Rembrandt’s Abraham’s Sacrifice of 1635 (Hermitage, Saint Petersburg). This free copy, which is now in Munich, has been attributed to different painters on stylistic grounds: Jan Lievens, Govert Flinck and Ferdinand Bol. These three names not only suggest different makers, but also different locations where the copy could have been made. In 1636, the year inscribed on the painting, Jan Lievens stayed in Antwerp, while Flinck and Bol worked in Amsterdam, whether or not in Rembrandt´s workshop. The copy has a signature, which reads Rembrandt. verandert. En overgeschildert. 1636 [Rembrandt. altered. And overpainted. 1636]. Although somewhat enigmatic, the signature suggests that the painting was made in Rembrandt’s workshop. This supposition is supported by the origin of the canvas on which the copy has been made. This canvas originates from the same bolt as two authentic paintings by Rembrandt, both dated 1635, the Minerva and Belshazzar’s Feast.1
For Vermeer, there is also an example of canvas research which supports a positive attribution. The painting Young Woman Seated at a Virginal (L36) was rejected as an autograph Vermeer by several art historians in the past. To the many arguments in favor of the acceptance of this painting as an authentic painting by Johannes Vermeer has recently been added that the canvas originates from the same bolt as the canvas of The Lacemaker (L29), of which the authorship is beyond doubt.2
In his research of Rembrandt’s canvases, Van de Wetering determined that with the majority of the two canvas pairs which have been identified as coming from the same bolt, the paintings concerned are companion pieces. He assumed that the canvases of these companion pieces were as a rule stretched and prepared in one and the same procedure. In all cases which Van de Wetering records, the companion pieces are portraits of husband and wife. This applies not only to Rembrandt, but also to Frans Hals. In almost all his companion portraits, of which the supports have been examined, the canvases also originated from the same bolt.3 This is useful in identifying portraits as companion pieces, when the knowledge concerning the link between two portraits has been lost in the course of time. This was the case with two portraits by Rembrandt of an unknown man and woman, now in Kassel and Vienna. Research of the canvas not only showed that the two fabrics came from the same bolt, but also that the two paintings were originally the same size.
In genre paintings, the identification of companion pieces is less obvious than in portraits. This becomes clear when one considers the differing views on this aspect in the paintings of Vermeer. His oeuvre includes two cases in which the question of whether we are dealing with companion pieces or not resurfaces time and again in art historical literature. In his recent monograph on Vermeer, Franits formulates a series of criteria for identifying pendants: both paintings invariably carry the same date (or dates very close in time to one another), have the same dimensions, complementary compositions, general stylistic affinities and related subject matter. According to Franits, The Astronomer (L28) of 1668 and The Geographer (L27) of 1669, which are often seen as pendants, do not fulfill all of these criteria. This is, according to Franits, contrary to the idea that the paintings are pendants, conceived as a pair from their inception. The two paintings were sold together at four different auctions during the eighteenth century, commencing in 1713, the earliest recorded reference to them. Noting their frequent pairing in eighteenth-century sales, many scholars believe these pictures are pendants. Franits rejects this idea and proposes a slightly different alternative. He considers it conceivable that a patron commissioned The Astronomer (L28) in 1668, and in the following year perhaps decided to request an additional work from Vermeer with a related subject, The Geographer (L27).4 Although this is an ingenious hypothesis, a recent finding seems to contradict Franits’ reconstruction of events and to advocate that both works were conceived simultaneously: both canvases originate from the same bolt.5 This supports the hypothesis that both paintings were intended as a pair from the beginning. This argument is put forward by Franits in support of the hypothesis that two other paintings by Vermeer are pendants, Young Woman Standing at a Virginal (L33) and Young Woman Seated at a Virginal (L34).6 The presumption of simultaneous creation as a pair of the latter two paintings is supported by an identical composition of the ground. It would be interesting to examine the ground layers of The Astronomer (L28) and The Geographer (L27) and to see if their ground layers also match.
1 E. van de Wetering, ‘The canvas support’, in: J. Bruijn et al., A Corpus of Rembrandt Paintings, 6 vols., Dordrecht/Boston/Lancaster 1982-2014, vol. 2 (1986), pp. 15-43, esp. pp. 28, 30.
2 L. Sheldon and N. Costaras, ‘Johannes Vermeer’s “Young Woman seated at a Virginal”’, The Burlington Magazine 148 (2006), pp. 89-97; W. Liedtke, C.R. Johnson Jr., and D.H. Johnson, ‘Canvas Matches in Vermeer: A Case Study in the Computer Analysis of Fabric Supports’, Metropolitan Museum Journal 47 (2012), pp. 101-108, esp. p. 102; W.A. Liedtke and A.K. Wheelock Jr., ‘Young Woman Seated at a Virginal’, in: A.K. Wheelock Jr. (ed.), The Leiden Collection Catalogue, http://www.theleidencollection.com/archive/ (date consulted: August 11, 2017).
3 E. Hendriks, K. Levy-Van Halm, in collaboration with J.R.J. van Asperen de Boer, Report concerning a preliminary technical investigation of paintings exhibited during the Frans Hals exhibition, held from May 11 to July 22, 1990 in the Frans Hals Museum, Haarlem, Haarlem 1991, pp. 4-11.
4 W. Franits, Vermeer, London/New York 2015, pp. 222-229.
5 C.R. Johnson Jr. and W.A. Sethares, ‘Canvas Weave Match Supports Designation of Vermeer's Geographer and Astronomer as a Pendant Pair’, in: Journal of Historians of Netherlandish Art 9 (2017), issue 1, http://jhna.org/index.php/vol-9-1-2017/348-johnson-sethares (DOI: 10.5092/jhna.2017.9.1.17).
6 Franits 2015 (note 4), p. 324 (note 23).