Oud Holland

Review of: ‘Study heads’ (2020)

November 2022


Review of: Nico van Hout, Corpus Rubenianum Ludwig Burchard part XX–study heads, Bert Schepers and Brecht Vanoppen (eds.), London/Turnhout [Harvey Miller], 2020. vol 1 text; vol 2 illustrations

The Corpus Rubenianum has been very productive of late. Nico van Hout’s Study heads is among the recent volumes to appear, and provides a very welcome, if ultimately non-definitive record of the production of tronies in the workshop of Rubens (1577-1640).This opening comment is not a criticism so much as an indication of the challenges that face experts in this important but neglected area of Rubens’ production, which is so instructive about his studio practice. Van Hout’s essay, catalogue and appendices provide the most important contribution on this subject to date. Providing a definitive catalogue is indeed the aim of the Corpus, but specific difficulties apply to the study of the tronies. Distinguishing originals from copies is always a problem. No part of Rubens’ work, however, has suffered more from a process of mutilation and overpainting. The most significant challenge addressed by Van Hout is identifying the artists who produced head studies when working alongside Rubens. These and other issues are very well navigated by the author who has examined a great many of the actual works and so reduced the number of judgements made on the basis of photographs. Van Hout’s opinions sometimes diverge from those who have gone before him, but he gives clear and confident proposals, especially but not restricted to matters of attribution.

One important advance in knowledge of the subject is that Van Hout has substantially extended the core group of known tronies. This has been done, first by the inclusion of lost works – that is panels whose existence can be deduced from the appearances of a particular head in multiple compositions. Importantly, a number of works not by Rubens are included. The series editor, the late Arnout Balis explained that, "A different category demands different editorial decisions, which in this case involve the inclusion, in the catalogue, of works which are not painted by Rubens’ own hand, and which cannot even claim to be of his 'invention', but which were nevertheless made on his instructions and under his direction" (p. 11). A great number of Rubensian head studies are in existence. Decisions about which of these to include is partly a matter of scrutinising the heads that appear in Rubens’ compositions, but it has also been aided by valuable research by Flemish scholars into artists from Rubens’ circle, whose work is less visible and harder to comprehend for those of us outside Flanders. That tronies were painted by at least one assistant is recorded in one of the few documents mentioning tronies, from which it emerges that a number were painted by Anthony van Dyck (1599-1640). The often-quoted list of Rubens’ works offered for sale after his death in 1640, records that the tronies were bundled together as a, "…parcel of Faces made after the life, vppon bord and Cloth as well by sr Peter Rubens as van dyke."2 It is particularly useful to be guided here by a scholar who has the eye of a conservator and has the courage to build on this mention of Van Dyck, and name others who demonstrably did, or may have created tronies for Rubens’ workshop. The evidence which Van Hout produces in this catalogue concerning Jacques  Jordaens (1593-1678), especially concerning the double head study in Nancy, Two head studies of an old woman (no. 77) (fig. 1), strengthens the argument he has advanced elsewhere that Jordaens and others played a creative role in Rubens’ workshop and did not merely do as they were told.3 One of the ways in which they actually contributed to the design process was in the painting of tronies.

Cover of Corpus Rubenianum Ludwig Burchard part XX–study heads

Middle right: fig. 1 Jacob Jordaens, Two head studies of an old woman with a headscarf, oil on panel, 58.5 x 65.5, cat. no. 77, Nancy, Musée des Beaux-Arts de Nancy (PD.13-1980)

So, what are tronies? Following in the footsteps of earlier masters of the Antwerp school such as Frans Floris (1519/20-1570), but also influenced by his experience in Italy, Rubens created – certainly from 1602 onwards – oil sketches of heads in expressive poses, representative of types in order to create a stock of heads for use in compositions. Some appear to have been used only once (for example Joseph of Arimathea on the right in the Antwerp Descent from the cross altarpiece, cat. 28), but others occur many times. The early Caravaggesque Youth with raised head, (cat. 1), appears at least nine times, and the two Head studies of a sleeping child (41) make a total of 15 known appearances. The word tronies means faces, or heads, for that is all that they were. Almost all were painted without indications of clothing or background. Rubens’ tronies were not for sale but were preparatory studies, designed for consultation and repeated studio use.4 Some secrecy surrounded them. In the one letter in which Rubens refers to a head study, his assistant the sculptor Lucas Faydherbe (1617-1697) is requested to bring a multiple tronie panel from which Rubens needed to work and he asks Lucas to keep it covered: "It would be a good idea to cover it with one or two new panels, so that it may not suffer on the way, or be seen."5

Oil sketches form an important and distinctive aspect of Rubens’ oeuvre. The most familiar are the compositional sketches on panel that he painted to show patrons who were commissioning a major work. Another type of sketch, drawn and painted on paper, and reproducing a painting by the master, was made as design for an engraver and these were often executed by a studio assistant (especially Anthony Van Dyck (1599-1641)) and retouched, if necessary, by Rubens. The tronies form a third and quite a separate category of sketch. Many were painted by Rubens himself but, bearing in mind what we know about designs for engravings, it should be no surprise that some workshop tronies were made by Van Dyck, Jordaens, and almost certainly others.

The tronies provided templates for the all-important heads. "The insertion of head studies in Rubens’ compositions can to some extent be thought of as a forerunner of the cut-and-paste technique carried out in modern image processing programmes" (p. 38). Having a varied stock of heads invented by the master, designed to take into account a variety of subjects and formats, was a means of controlling quality and maintaining uniform studio production. The tronies are mainly of middle-aged or older men, of robust-looking, saintly types. Curly hair and vigorous grey beards abound (fig. 3). There are some younger men, but only a handful of women, and even fewer children. All adopt a distinctive, active pose, looking up, or down, or to one side – only rarely do they look towards the viewer – to facilitate their placing within a lively group of interacting figures. "Their facial expressions tend to be neutral …for they were designed to be used in different contexts" (p. 71). The tronies are of central importance to Rubens’ compositions because they provided the structure onto which character and expression were grafted in the finishing of the work. It is mainly through the expressions on the faces of characters that we understand the meaning of any work.

The versions of the heads that appear in finished compositions correspond in most cases very closely to the tronies: in type, in pose, and in facial features such as bone structure, marks on skin, hair, beard etc. Occasional variations are made to the hair; the angle of a pose might differ from the prototype, but only slightly. Van Hout is rightly always cautious when proposing that a similar head represents a modified version of a tronie. Sometimes even a small difference can be better explained by positing the existence of a lost tronie. There are numerous pairs of very similar heads, for example (56), (66, 67), (68, 69), or Van Dyck’s two versions of the bearded man grimly averting his gaze (100) (fig. 4). The point of providing closely related heads was to create prototypes that needed very little, if any, adjustment. The similarity of the pairs is deliberate, meaning that each panel provided four possibilities: the head to the left, the one to the right, and the mirror image of each.

Early examples, including at least one demonstrably painted in Italy (1, possibly also 6, 9), are in oil on paper, but on his return to Antwerp Rubens started to build a library of types painted on oak panels. The principal period of tronie production corresponds to the years 1609-1620 when the workshop’s extraordinary production required the assistance of many able painters. Some managed to make the leap from painting backgrounds, or small elements, to designing whole compositions on the basis of tronie heads. A familiar example of a painting designed in this way and intended, it seems, for the open market, is the St. Ambrose altarpiece in Vienna (c.1618-1619). Van Hout names the Ambrose painting – with eleven heads – in the entries for eight tronies (mentioning also a drawing of the head of the Farnese Hercules): 10a, 12a (left), 36b, 56 (left), 61, 64, 85, 102b; to which 11a should be added. The leading role that Van Dyck played in this painting is documented by the second, better, version of the composition that he painted (London, National Gallery), and in the inscription by Willem Panneels (c. 1590-1634), naming Van Dyck on one of his copies of preparatory drawings for the work.6

There is hardly any documentation about the tronies and what survives is not helpful in narrowing down the date range of Rubens’ activity. Dates, however, can be established using the internal documentation of workshop commissions as well as a small number of ‘compilation drawings’, usefully included by Van Hout as a supplement to the main catalogue, describing them as "aides-mémoires when choosing and inserting tronies and as points of departure when designing new compositions" (p. 60). Besides the mention in the Spécification, the only other document which helps understand Rubens’ use of tronies is the letter of 1638 already mentioned in which the painter wrote from Het Steen requesting: "…a panel on which there are three heads in life-size, painted by my own hand, namely: one of a furious soldier with a black cap on his head, one of a man crying, and one laughing."7 We learn that besides directing assistants to insert specific tronies into workshop compositions, Rubens worked this way himself, and was inserting his own pre-designed heads into paintings even when working alone many years after the peak period of workshop production.

Rubens’ letter concerns a panel with three heads, and Van Hout’s catalogue confirms that there were originally many such multiple studies, usually with two or three different poses of the same head painted, economically, close together. (One famous example is the painting in Brussels with Four head studies of a young African man, no. 53.) After Rubens’ death we can identify a period in which artists close to Rubens acquired tronies, as mementos perhaps, but also using them in their own work and passing on knowledge of the process to their own pupils. Eventually the study heads found their way onto the market, which sometimes valued them for what they were. Some individual heads feature in paintings of noble collections.8 The multiple panels fared less well, however, and many were mutilated in order to make them more saleable. The head studies were objects with a studio function, and multiple poses of the same man made little sense to those who then wanted to buy a Rubens. The panels were made more valuable by sawing the heads apart and extending them to restore balance and allow for clothing and attributes. Van Hout gives the evidence of who owned and therefore might have been responsible for the transformation of heads into subject pictures, found in the inventories of artists from Rubens’ circle. So we know, for example, that the painter Erasmus Quellinus II (1607-1678) owned 20 Rubens tronies, but he also had 7 by Van Dyck. Jan Wildens (1586-1653) owned 7 by Rubens and 16 by Van Dyck. Another artist who acquired a number of tronies (nos. 35b, 48, 49, 64, 66, 82 and 108) was Jan Boeckhorst (c. 1603-1668), a pupil of Jordaens and Rubens, whom Van Hout identifies as the painter responsible for additions to several tronies, including the Bearded old man, looking up (82), today in Frankfurt. In its present state, this is a Rubensian painting (84 x 68 cm.) of King David singing the psalms. The head is a fragment (61.5 x 48.2 cm.) that was once half of a double tronie panel, and which an artist has placed, like a precious jewel, into a new setting, by adding sections of panel. It is interesting to note that, in a field in which doubts persist about which of several versions of a tronie is Rubens’ original, finding a panel that has been sawn apart and over-painted by Boeckhorst is the closest thing we have to a guarantee of authenticity.

Mercifully a good number of multiple panels survived the ravages of Boeckhorst. Van Hout has rightly attempted to reconstruct these works, and he prefers to attribute to Rubens, or name as the original, a head which appears to have been separated from its companions: a good principle. Several former multiple tronie panels have been reconstructed and given paired numbers. In several cases (10a/10b; 11a/11b; 34a/34b; 36a/36b; 44a/44b; 52a-/52b; 75a/75b; 96a/96b; 102a/102b) there is good evidence to reconstruct the panels, in the form of visible saw-marks, matching grain of left and right halves, as well as evidence of relative positioning provided by copy drawings and the compilations (A-P). In some cases, where understandably it has not been possible to compare panels side-by-side, this is speculative (12a/12b; 35a/35b).

Left: fig. 3 Peter Paul Rubens, Head of a bearded man, pen and brown ink, heightened with white, 100 x 74 mm., Hunterian Art Gallery, University of Glasgow, GLAHA:54302
Middle: fig. 4 Anthony van Dyck, Two head studies of a bearded man, looking down, oil on canvas, 45 x 67 cm., private collection., cat. no. 100. Van Dyck’s powerful double tronie canvas played an important role in designing the Miraculous Draft of Fishes altarpiece in Mechelen (1618-1619), and confirms the long-held belief that the artist played a major role in that commission. A careful look at the related oil sketch in the National Gallery London reveals that Rubens revisited the tronies used in 1618 when redesigning the image for the engraver Schelte à Bolswert c.1638. This Van Dyck canvas emerges, even more clearly from the London sketch than from the altarpiece, as playing an important role in both works. The head of the oarsman in Christ’s boat, high up in the centre of the composition, corresponds to the left-hand head. In both the altarpiece and the London sketch, the left-hand head was also used for the man in the front boat who reaches forward to drag the net ashore. The right-hand head of Cat. 100 was used, in mirror image, for the oarsman standing in the rear boat, grimly concentrating on his task.

Right: fig. 5 Peter Paul Rubens, The miraculous draught of fishes, 1618-1619, extended c.1638, black chalk, pen and oil on paper, stuck on canvas, 55 x 85 cm., National Gallery, London NG680. Although it is not certain when it was begun, this work was most likely completed c.1638 as design for the engraving by Schelte à Bolswert, whose collaboration with Rubens can be dated from that year. The central, vertical portion of the image matches the altarpiece made c.1618-1619 for the Mechelen Fishermen’s Guild (Bulckens Ministry of Christ, no. 18).

One way of assessing this catalogue is to count the total number of tronies. Van Hout catalogues 129. Some years ago I began a catalogue of these works by recording surviving examples, arriving at about 50. However, besides listing known works, Van Hout has pieced together his own observations and those of others in order to make entries for heads that appear more than once in Rubens’ compositions, but for which no prototype survives. In this way he has added forty-plus entries for works that may one day resurface (including 2-5, 11a, 14, 18-26, 29, 36a, 39, 40, 43, 44a-b, 51, 58, 61, 63, 65, 71, 73, 75a, 84, 86-9, 91-2, 105, 112-115, 117). Whether this overall figure is too high or too low cannot be said. A quite specific number emerges in a 1997 article by Martin Eidelberg. In researching drawn "Rubens’" heads copied by Watteau, he focused attention on an album containing 94 heads, which was traceable for a time alongside Rubens’ famous lost theoretical notebook, now tantalisingly lost.9 As Van Hout states (pp. 60-61), "the rationale for the existence of the “heads” album lies in Rubens’ workshop practice. The vast amount of study material made it essential to have a good indexing and classification system."

I began by saying that the catalogue is not definitive, but I do believe that it will prove ground-breaking in making sense of the creative contribution of Rubens’ assistants. The main group of tronies not by Rubens are by Van Dyck. There is one very significant work by Jordaens, and other known artists lurk unnamed behind the term ‘Rubens workshop’. Future technical research may help to name the painter of more, or restore some to Rubens. The author states in his Preface (p. 13) that the catalogue brings together, "all the painted head studies (tronies in old Dutch) that are known for certain to have been used in a Rubens composition" (p. 13). But there are 7 entries (70, 116, 119, 120, 122, 127, 129) for which no uses in Rubens compositions are mentioned. Cat. 70 was clearly included because it was identified as autograph, by Burchard. But the presence of the others, four of which are by or attributed to Van Dyck (116, 119, 120, 122), merits explanation. 

Catalogue no. 77 is a panel, today in Nancy, showing two poses of the same beaky toothless Old woman with a headscarf.10 It is in fact the tronie panel used more often in Rubens’ compositions than any other, so it will be a surprise to many to find that this work was painted by Jordaens. Van Hout lists no fewer than 15 Rubens compositions, dating between approximately 1617 and 1624, in which one or other of the two heads of this impressive old woman appear. This brings me to a point about attributions, or about how they are expressed in this catalogue. It seems to me that if a particular tronie does not appear in Rubens’ work, there is reason to question its attribution – other criteria must then be brought in to justify the attribution. Similarly an apparent original which appears more than once in workshop compositions has a good claim to be by Rubens, as with this magnificent double trony. This works both ways: even though I have not had an opportunity to see cat. 77 for myself, I am convinced that Van Hout is right in identifying as by Jordaens the tronie which was given the most outings in compositions by Rubens. But it is characteristic of the entry for this work that the author does not methodically lay out the steps of an argument for his attribution. He lists as No. 77 (copy 5), the red chalk drawing in Copenhagen which shows the right-hand head. Since this drawing is accepted as by Jordaens, its relationship should surely be explained as supporting evidence and it should not be described (p. 58), confusingly, as a copy.

A number of works by Van Dyck are included at the end of the catalogue, some of which (95, 96a, 99, 100, 101, 126) were demonstrably used in Rubens compositions. Others appear in both Rubens’ and Van Dyck’s (95, 100, 101). Among the works that follow, by Van Dyck (116, 119, 120, 123, 126, 129), attributed to Van Dyck (122) and by an unidentified hand (128) we have a group of tronies which according to the author’s prescription in the preface (p. 13) ought not to be included because they are not, "known for certain to have been used a Rubens composition". Among those for which no Rubens use is listed is a fine Rubensian head study (127) which is catalogued as: "Attributed to Artus Wolffort". Not having been used in Rubens’ work, it is by definition the design and the work of an assistant or follower. However, no criteria are given for this attribution and the reader is left with the disappointing feeling that she/he has been excluded from a conversation. One advantage of including these works is that others can build on the selection of head studies, which certainly achieves a focus on the work of Rubens’ assistants, much of it impressive. I found myself drawn to confront the head studies by Van Dyck in a way I had not before, looking especially at the Miraculous draught of fishes altarpiece in Mechelen (fig. 4), which was one of the compositions created using head studies, and is a work in which the hand of Van Dyck has been detected. In the Corpus Volume 5 (2), The ministry of Christ that treats the work, Bulckens stated that there, "are few motifs in the work, like heads, that recur elsewhere in Rubens’ oeuvre". I can find appearances in the altarpiece of several heads included by Van Hout and they confirm the close involvement of Van Dyck. Using the images of the Van Dyck head studies now available in these volumes, it is particularly interesting to look at the London oil sketch which Rubens presumably made to update and improve the composition when commissioning the engraving by Schelte à Bolswert (c. 1586-1659). This would have been in about 1638, the year that appears on one of the series of Bolswert’s Small landscapes (Hollstein 316).11

Some of the heads in the London sketch correspond very closely to tronie heads (as well as the corresponding heads in the altarpiece) and have a vividness that suggests that Rubens was consulting his stock of head studies afresh. The head of the central standing helmsman is very close to the left-hand head in the Van Dyck double tronie (cat. 100) and was probably used as model for the figure at the apex of the altarpiece. The same head appears to have been used in the London sketch, with head angled downwards slightly, for the central figure stretching his arms forward to drag the net ashore. Even though the angle of the head is slightly different, this probably indicates that this extraordinary double head study by Van Dyck was indeed the tronie used twenty years earlier for the matching figure in the altarpiece. Confirmation can arguably be found in the head of the standing figure with oar in the second boat. His features and downcast expression closely resemble those of the right-hand head in (100) in mirror image.

I have for the sake of convenience sometimes put elements of Rubens’ procedure into my own words but the essay accompanying the catalogue addresses, I believe, every question on this subject that needs answering. The catalogue entries, also include a record of technical investigations, images of which are included as figure illustrations. Following the Corpus’ standard format, there is a subject index. This seems superfluous since there is really only one subject: the head. Perhaps, not surprisingly, users may find that they most often begin a new investigation by consulting volume two – for the extremely valuable sequence of illustrations of the heads, now available in one place, for the first time. Rather as Rubens’ assistants may have done, using the lost ‘heads’ album.

Peter Black
Honorary Curator of Prints
Hunterian Museum and Art Gallery, University of Glasgow

1 I am very grateful to Martin Royalton-Kisch, Tico Seifert, Isabelle van Tichelen and Simon Turner for advice on dates and attributions during a time when I had no access to books.

2 For the Spécification, see: J. Muller, Rubens. The artist as collector, Princeton 1989, p. 145.

3 Notably the article in Rubensbulletin 2012 about Jordaens’s role in the Lille Descent from the Cross. As an aside, I would say that it was not every assistant’s experience that Rubens allowed creative interventions. Most know how the collaboration with Lucas Vorsterman (1595-1675) ended. But then, Rubens’ creative partnerships with Van Dyck and Jordaens are surely evidence of the engraver’s madness.

4 Confusingly, tronie is also the word used for a genre of seventeenth-century painting common in the northern Netherlands, and made for the open market. Those very different paintings sometimes show monumental figures, which are not portraits of individuals but representative of types, or examples of particular emotions.

5 R. Magurn, The letters of Peter Paul Rubens, Cambridge 1955, no. 244, p. 410. The autograph notes made by Willem Panneels on his copies of Rubens’ drawings and tronie panels also show that he knew that this material was what we might call ‘copyright’, or ‘trade secrets’ of the artist. See: J. Garff and E. de la Fuente Pedersen, Rubens cantoor. The drawings of Willem Panneels, 2 vols, Copenhagen 1988. D. Hirschfelder, ‘Portrait or character head: the term "Tronie" and its meaning in the seventeenth century’, in E. Van der Wetering (ed.), Mystery of the Young Rembrandt, Munich 2001, pp. 82-90. D. Hirschfelder, Tronie und Porträt in der niederländischen Malerei des 17. Jahrhunderts, Berlin 2008.

6 J. Garff and E. de la Fuente Pedersen 1988 (note 5), vol. 1 p. 9, no. 25.

7 R. Magurn, The letters of Peter Paul Rubens 1955 (note 5), no. 244, p. 410.

8 (Van Hout quotes the example of Storffer’s views (1720-1733) of the Stallburg Picture Gallery in Vienna.)

9 The album is cat. 840 in ‘Description Sommaire des Desseins des Grands Maistres d’Italie, Des Pays-Bas, et de France, Du Cabinet de feu M. Crozat’, in the section headed ‘Etudes de Têtes de Rubens’. Apparently: ‘Ces quatre-vingt-quatorze Têtes formaient ci-devant un volume que M. de Piles avait apporté de Flandres, & qu’il conservait précieusement. L’on croit qu’il l’avait trouvé dans le même endroit que ce Manuscrit singulier du sieur Rubens, qui ayant passé entre les mains du sieur Boule a péri, avec tant d’autres curiosités, dans l’incendie de sa maison’.

10 This panel by Jordaens holds the record for the tronie that appears the largest number of times in compositions by Rubens. Van Hout lists 15 compositions, dating between approximately 1617 and 1624, in which one or other of the two the heads of this impressive old woman appear.

11 Some additional information can be gleaned from a tronie fragment by Rubens (fig. 3): Head of a bearded man, pen and brown ink, heightened with white, 100 x 74 mm., Hunterian Art Gallery, University of Glasgow GLAHA:54302. Van Hout mentions this fragment, torn from a compilation drawing, (p. 59, note 179), but it merits inclusion in the catalogue, among the compilation drawings. Like those drawings, it records a lost tronie, which was possibly used for the head of the bearded man in red bottom left, in the centre panel of the Miraculous draught of fishes altarpiece in Mechelen (1618-1619, Bulckens no. 18).
Peter Black, ‘Review of: Corpus Rubenianum Ludwig Burchard part XX–study heads’, Oud Holland Reviews, November 2022.