Drawn by the Brush: Oil Sketches by Peter Paul Rubens
About the Exhibition Online Exhibition Timeline Catalog Essays Teacher Resources Credits

Peter Paul Rubens
The Triumphant Entry of Constantine into Rome (detail), 1622
Oil on panel, 48.6 x 64.5 cm
Indianapolis Museum of Art, The Clowes Collection, inv. IMA2001.237

by Peter C. Sutton

Two Hundred Years of Collecting Rubens's Oil Sketches

By Marjorie E. Wieseman

by Nico van Hout



Catalog Essays

by Nico van Hout

In many museums Rubens's oil sketches occupy a glittering place of honor. And so it should be. Are not the master's virtuosity and his famously sure and fluent brushwork expressed most clearly in these works? Moreover, the originality of the oil sketches is beyond dispute (or so one assumes), while it is increasingly becoming a cliché that Rubens's large-scale works were to a significant extent painted by a ghostly yet crowded atelier of assistants.

However, oil sketches belong only to the preparatory conceptualizing that precedes the final painting. They are inherently a means and not an end. Perhaps Rubens would find it incomprehensible that the public today values his unfinished preparatory works more than his elaborate history paintings. There is no doubt, however, that Rubens's final goal was the completed painting, which was the result of an intense wrestling with form, color, and the application of the brush. Sauntering through the galleries of museums or leafing through publications with reproductions of Rubens's oil sketches, one is constantly struck by how different they look one from another. Some sketches consist of barely more than a few suggestive dark outlines and highlights, which Rubens applied ever so nimbly and efficiently to the streaky imprimatura layer. This buffer layer makes the lime-chalk ground of the panel less absorbent, so that the colors retain their brilliance. The imprimatura layer is applied with a quick zigzagging movement and functions in the sketch as a middle tone (a little like the blue, rose, and green ground layers of Renaissance drawings).1 It is only at a certain distance and with sharp eyes that the viewer is conscious of these effects in the picture: it is seen, among other places, in Rubens's interpretation of the Crowning of Mary by Christ in the company of God the Father and the Holy Ghost, one of his series of sketches for the ceiling decorations of Antwerp's Jesuit church (Museum Boijmans Van Beuningen, Rotterdam).

The Rubens oil sketch in the Kimbell Art Museum in Fort Worth (The Duke of Buckingham) has an entirely different appearance. The duke of Buckingham and his horse are worked out in all details. The sketch is a scrupulously faithful record of the large portrait that the English politician had ordered of himself. Nothing in this scene is left to the imagination. The drapery of his fluttering red mantle is fully elaborated, while the reddish brown color of his horse and the ocher-yellow skin tone of Neptune are indicated in the foreground. Only a few dimly colored zones permit the streaky imprimatura layer to show through.

The heterogeneous character of the oil sketches thus demands a scientific explanation. Among Rubens scholars, Julius Held has most concerned himself with this question. In case after case he examined how the preparatory material related to the end result, namely the completed painting. Held did not arrive at a strict division of categories. Rubens's working methods resist reduction into a few formulas. Rather, they answer to an internal logic, a driving system that leads to the end result. The extent and degree of difficulty of his inventions demand a clear description of the way they were made.

In the painting tradition of the fifteenth and sixteenth centuries the invention of a composition preceded the act of painting. The act of painting itself was considered to be nothing more than a careful copying of the design with the aid of paint. Artists then applied very thin and transparent layers of paint and sought to alter as little as possible the original concept during the execution of the painting. Any drastic alteration would be disturbingly visible through the paint layers. Consequently, artists placed great stock in the careful preparation of the initial conception of the composition. 2 In De pictura (1453) by Leon Battista Alberti we read that it is preferable to make corrections in one's imagination rather than on the panel, since it is not necessary to draw back the bow when the arrow is not pointed at the target. 3

Venetian artists like Titian and Tintoretto departed from the traditional system. With their opaque oil painting technique, they created their compositions during the act of painting, correcting their errors directly on the canvas. Rubens was familiar with both concepts. As a result he left a great quantity of preparatory oils sketches and drawings that show many large and small changes to his paintings, which one can study, for example, in x-ray photos, infrared images, and normal light.

In conceiving a painting Rubens in many cases followed a standard procedure, which began with the studying of the biblical, mythological, or historical subject that he was representing. He was supported in this research by his knowledge of classical literature and the allegorical language of imagery.4 Above all, Rubens, more than other artists of his day, must have had an astounding visual memory. His important sources of inspiration included antique sculpture and the inventions of the great masters of the Italian Renaissance, which he assimilated into his work. Besides his memory, Rubens also consulted drawn copies he himself had made or drawn reproductions that he purchased and extensively retouched and corrected. 5 From this gathered universe of forms, poses, and expressions he was able to give form to the most diverse subjects.

Then came the moment when he committed his first mental stirrings (pensieri) to paper. Just like other artists, Rubens probably made use of sketchbooks, although some compositions could also have been worked out on loose sheets. 6 The first tentative scribblings, what the Dutch called crabbelinge and the Italians described as scizzo, looked a little like the wandering lines that emerge on a notepad during the course of a long telephone conversation. In all probability a great number of such initial sketches did not survive the test of time. However, their number must have been considerable. In these first, loose pen strokes Rubens began in grosso modo the outlines of the figures that he would draw with pen or chalk and sometimes with graphite. 7 The first, summary contours were almost always changed soon after Rubens executed them. On top of an existing sketch he drew and worked up or elaborated several figures or groups and repeated certain elements of his composition, seeking alternative poses for his figures. Order was created out of the resulting chaos as he washed parts with his brush and diluted ink, so that the emerging forms became legible once more.

Thereafter Rubens made the first synthesis of the various crabbelingen in the form of a summary oil sketch on panel, or bozzetto. In this type of sketch in a small format the artist tried to find a solution for the treatment of light in his subject. Consequently, the choice of colors in bozzetti are restricted mostly to earth colors and white highlights. A second category of oil sketch, modelli, gives a more complete picture of the end result. Often, these modelli are actually bozzetti, that are more fully developed in terms of color and details. The creative process is here crystallized into a blueprint, a plan that is essential to the realization of the large-scale composition. Often the most detailed sketches were destined for tapestry weavers, sculptors, and jewelers who did not work in Rubens's workshop. Presumably Rubens determined the degree of finish of his oil sketches by his estimation of the abilities of the person or team who was to carry out his working document. It stands to reason that highly trained and talented assistants needed fewer instructions to complete a painting than collaborators who worked more independently. In this case, a bozzetto must have been more than sufficient. The differences between Rubens's oil sketches thus can be explained in part by the medium in which the subject would be realized and the qualifications of those who would carry out the project.

Finally, Rubens made detailed drawings of complicated figural positions and anatomical details that would be of use in the execution of the painting on a large scale. For these drawings naar het leven (from life) he posed models. In addition, Rubens used a number of tronies, or character head studies of personages (see Head of a Youth, Head of a Young Warrior, and Head of a Negro), that turn up in the most diverse compositions.

With the modello, the drawings, and possibly several tronies as a guide, one or more assistants began with the underpainting, or doodverf (literally, "dead coloring"), of the eventual painting. 8 The small subject was converted to a large scale; to ensure the accurate transcription of the composition, a grid in white chalk might be applied. The doodverf stage is a pale prefiguration of the final product, painted with pigments of lesser quality and less coloristic brilliance. The doodverf stage combines prefatory but well-defined ideas with the option for the artist to further refine the composition, or herdoodverven (re-dead coloring). Hence the oil sketch is not the final stage of the creative process.

After the dead coloring and the re-dead coloring comes the working up (opschilderen) of the painting and the retouching (retoucheren). The degree to which Rubens was involved in these later stages depended on the delivered work (inferior work had to be improved by retouching), the prestige of the commission, or the size of the payment that the artist would receive for the work.

One might assume that Rubens's standard procedure followed approximately this progression. But sometimes ad hoc solutions occurred to the painter while he was working, or he found a way to simplify the process whereby the usual intermediary stages in the development of a composition were passed over. This variable method naturally had implications for the function that the oil sketch assumed within the preparatory material for a particular painting. This elastic system affected the function of any creative step leading to a Rubens composition, including the oil sketches.

This early trial-and-error stage in the development of a composition was not restricted to works on paper. In working up some individual compositions Rubens chose to let his thoughts run free, taking their own course on the panel he had designated for his oil sketch (see The Martyrdom of St. Paul). 9 From the gossamer but illegible web of chalk lines, which often are detectable with the naked eye beneath the paint film, the ultimate composition would be spun out. In working these designs out, some embryonic chalk lines were ignored, while others were copied, strengthened, or newly defined with paint and brush.

It is striking that Rubens worked in this fashion mostly later in his career, in designing extensive series. He may have done this to save time and to get a more complete overview of the various components of the project. One can mention in this context the oil sketches for the tapestry series The Triumph of the Eucharist (1625-26) and The Life of Achilles (1630-32), as well as the sketches for The Triumphal Entry of Cardinal Infante Ferdinand into Antwerp (1634-35). In the oil sketches for this last commission, Rubens began by indicating a central vertical axis in black chalk, undoubtedly to get a better sense for the symmetry of his architectural design. A significant number of these tentative scribbles can be discerned with the naked eye beneath the paint layer on the oil sketches for (among others) the front and back façades of the Arch of the Mint in Antwerp. This is also the case with the other oil sketches that Rubens made for the project, and which are now in the State Hermitage Museum in St. Petersburg. A more complete picture of the underdrawing in both Antwerp sketches was obtained with the help of infrared reflectography. The visual evidence suggests that stylistically, the underdrawing on the oil sketches exhibits the same characteristics as Rubens's quick sketches on paper: made with swift movements of the wrist, they set out the most important lines of the composition. The underdrawings seem to have functioned for the artist primarily as an aide-mémoire or stenographic notation. They do not describe exact forms but instead roughly indicate the general placement of a figure, a niche, or a coat of arms within the design. In studying the underdrawing on sketches for the Achilles series and the sketch for the back side of the Arch of the Mint under the stereomicroscope, it appears that the underdrawing was applied on top of the streaky imprimatura. 10

In addition to using chalk, on rare occasions Rubens formulated his initial thoughts for a composition in a brush drawing on the panel. 11 In a handful of oil sketches, such as the Lion Hunt in London and Munich, Rubens worked up the salient details of the rearing horse and the falling rider more completely than the other—secondary—figures in the composition, which remain at the stage of a brush drawing. 12

In a number of instances Rubens appears to have barely prepared his paintings at all. Henry IV at the Battle at Ivry, at the Rubenshuis, Antwerp, is a fragment of one of the paintings from the uncompleted series devoted to the life of Henry IV. Apparently, apart from two small pencil drawings on the verso of another oil sketch, Rubens did not resort to any drawing or developed oil sketch. In this case the creative process took place directly on the canvas. Clear evidence of this is in the tentative lines and amorphous forms of this painting, which offer valuable information for the study of Rubens's painting technique. 13

Even within designs for the same series the execution of the individual sketches can be quite different. That is certainly the case with The Death of Hector, for example. 14 In comparison with the other designs for the Life of Achilles series (all now in the collection of the Museum Boijmans Van Beuningen, with the exception of Studies for a Figure in Larder), which are all rather more finished, this sketch shows an earlier stage of development, in which the painter strengthened or altered the course of his initial chalk underdrawing by means of fine brushstrokes in a thin, transparent brown medium, and added further details with the brush in dark gray, brown, and black. A few economical highlights indicate reflections in the armor. In Rubens's choice of technique for this sketch, drawing (disegno) takes precedence over color (colore). The red of Achilles' mantle and the yellow of Hector's are suggested with just a light wash of color. The decorative forms at the lowest edge of the composition are described with a minimum of brushstrokes: it seems as if this oil sketch was abandoned at the dead coloring stage.

Moreover, it is quite striking that in this same Achilles series, apart from differences between the individual sketches in terms of quality and degree of finish, the quality of brushwork can fluctuate within a single sketch. This is especially clear, for example, when we compare the herms that flank the compositions in the various oil sketches. Some of these figures are weakly and hesitantly painted (as in Thetis Dipping the Infant Achilles into the River Styx, Achilles Recognized among the Daughters of Lycomedes, Briseis Returned to Achilles, Thetis Receiving the Arms of Achilles from Hephaestus), while in other herms one can immediately recognize the sure and judicious brushwork of the master himself (Achilles Educated by the Centaur Chiron, The Wrath of Achilles, The Death of Hector, and The Death of Achilles).

It is not a simple matter to arrive at an explanation for these qualitative differences. The uneven state of preservation of the sketches might be one possible reason. Rubens painted his oil sketches quite thinly, and they are thus extremely vulnerable to solvents. That is certainly the case for the earth pigments that Rubens used to give form to his figures and model them with light and shadow. It is precisely these transparent paint layers that over time have been harshly removed during careless restorations. These damaged areas were frequently repainted. In some oil sketches the streaky imprimatura has been thoroughly scrubbed, with the paradoxical result that even the strongest highlights come across as darker than their bleached surroundings.

Another explanation may lie in the possibility that Rubens relied on his assistants more in the creative process than has hitherto been assumed. It is certain that Rubens was responsible for the final production of the series of designs for paintings and tapestries that were realized by his studio. But must this automatically also mean that the master designed each composition for a series of related images with his own hand? Did Rubens delegate work simply and solely when he desired copies to be made? Or did he sometimes give his most talented assistants—those who could live up to his artistic expectations—the task of coming up with their own proposal for a composition?

Apart from that, in the case of tapestry designs, it is highly likely that the patron or the tapestry dealer might initially have envisioned a smaller series and the order was subsequently enlarged with additional pieces. That is perhaps the case with the thirteen designs for the Life of Constantine, of which twelve were woven in Paris in the shop of Marc Comans and François van der Plancken (French, de la Planche). Seven of the tapestries were given by Louis XIII to Cardinal Francesco Barberini as a farewell gift when he left Paris in 1625. The series of oil sketches for the project can be divided into two groups. The first group consists of broadly conceived compositions with numerous figures and battle scenes. 15 The second group consists of rather simple, static compositions with fewer figures, painted on smaller panels with squarer dimensions. 16 The beautifully executed Triumphant Entry of Constantine into Rome belongs to the first group of sketches. In comparison with the brushwork in this sketch, the handling of paint in The Labarum, a sketch from the second group, is much less lively. The paint is applied evenly and there is hardly any transparency to speak of. The Constantine series was designed at the precise moment that Rubens's atelier was working at top speed to complete the paintings for the Marie de Médicis cycle before the specified deadline. It is therefore not unlikely, given the immense amount of work, that Rubens was forced to entrust some of the designs for secondary projects—even for paintings or tapestries—to his assistants.

Held observed that small, irregular marks, scratched with the end of the brush into the wet paint, can be found in a number of oil sketches by Rubens: a phenomenon encountered earlier in the work of (among others) Quentin Metsys (The Lamentation, Koninklijk Museum voor Schone Kunsten, Antwerp) or Pieter Bruegel the Elder (The Adoration of the Magi, The National Gallery, London); and later in works by the young Rembrandt (here we think of the wiry lines incised in the dark mop of hair in his Self-Portrait in the Rijksmuseum, Amsterdam). 17 On rare occasions Rubens also experimented with similarly incised scratches. In St. George and the Dragon (Museo del Prado, Madrid), for example, he used the wooden end of the brush to scratch lines in the wet paint to suggest the white hairs in the horse's tail. The scratched traces on the oil sketches, on the other hand, appear to be meaninglessly and inadvertently introduced; they are more mechanical accidents, the results of nonchalance, rather than painterly devices to create an effect of three-dimensionality. Until now it has been somewhat overlooked that these mysterious little lines are found mostly along the right edge of the sketch, mostly near the top, sometimes near the middle, and only rarely near the lower right edge. We find them again in a number of larger panels with a standard format of about 65 by 50 centimeters, 18 and on a few panels that were conceivably sawn down to a smaller format at some later date. 19 On the sketch for the Triumphal Chariot of Kalloo in Antwerp, the viewer can also notice identical scratches in the middle of the upper portion of the panel, mostly to the right, or above and to the right, of a more extensively painted area. The marks probably occurred as Rubens in the course of painting his sketches made use of a maulstick (or a long-handled paintbrush), which he held in his left hand and on which he rested his right hand as he painted. Fijnschilders used sticks that were furnished at the tip with a soft cushion to prevent damage to their diminutive paintings. Because the oil sketches were, after all, just preparatory material, perhaps Rubens was content with using a maulstick with a blunt end. The contact and movement of the instrument against the panel could thus have caused the scratches. Ordinarily one would not associate the use of a maulstick with the broad and free brushwork of Rubens, the purveyor par excellence of monumental works. But to guide a hand that had to prepare these complex compositions in a reduced format, a maulstick would certainly not have been superfluous.


  Bruce Museum of Arts and Science :: 1 Museum Drive, Greenwich, Connecticut :: 203.869.0376
  UC Berkeley Art Museum and Pacific Film Archive :: 2575 Bancroft Way, Berkeley, California :: 510.642.0808
  Cincinnati Art Museum :: 953 Eden Park Drive, Cincinnati, Ohio :: 513.721.ARTS

  Questions or comments about this site? :: Copyright Information