Frank Zöllner
Rubens Reworks Leonardo:
`The Fight for the Standard'
I The Problem of the copies
LEONARDO DA VINCI`S most important commission as a painter was the `Battle of Anghiari', an over-life-size mural painting for the `Sala del Gran Consiglio' in the Palazzo Vecchio in Florence, depicting the battle between the Florentine and Milanese troops in 1440. Leonardo never finished this commission for the Florentine Republic but he seems at least to have executed a cartoon and a wall painting of a central group, the so-called `Fight for the Standard', which included four horsemen fighting for the Milanese standard and three more soldiers fighting or crouching on the ground. The mural itself must have been destroyed between 1540, when the Medici moved into the Palace, and the early sixties of the sixteenth century, when Vasari finally covered the walls of the former Sala del Consiglio with his own paintings. Roughly within the same period the cartoon also perished. After the loss of both the cartoon and the wall painting, our knowledge of Leonardo's original creation is based on his preliminary sketches, on a number of copies of the `Fight for the Standard' by contemporary artists and on some early descriptions.
The major problem for a modern reconstruction of Leonardo's `Battle of Anghiari' derives from the artistic quality and reliability of the early sixteenth-century copies. An aesthetically acceptable copy, such as the sketch by Raphael, lacks most details (fig. 4), while the more detailed interpretations often are of questionable reliability or inferior artistic quality. Both superior artistic quality and a satisfactory amount of detail are found only in a large drawing in the Louvre (fig. 1), traditionally attributed to Rubens, who, however, having been born in 1577, could have seen neither the cartoon nor the wall painting. But the Leonardesque quality of the Louvre drawing and of two of its early copies (figs. 1-3) has often been acknowledged, and consequently some scholars have credited this quality to the artistic empathy of Rubens, who is thought to have understood Leonardo by a kind of intuition. Others have tried to explain the Leonardesque quality of the Louvre drawing or its early copies by assuming that Rubens had known either Leonardo's original cartoon, a very exact copy of the original composition or Leonardo's so-called `trial panel' (which I shall discuss below). However, Rubens could not have seen the cartoon, no exact copy has ever been identified, and the importance of the `trial panel' can hardly be assessed, since we lack precise information about its appearance. Hence, today most scholars consider the Louvre drawing to be the synthesis of a combination of early copies and descriptions, and according to a general consensus the drawing traditionally attributed to Rubens and its early copies can hardly have any value for our knowledge of Leonardo's original composition.
In the following paper, I reconsider the importance of the Louvre drawing and of its early copies (figs. 2-3) for our understanding of Leonardo's ideas for the `Fight for the Standard', because the attribution to Rubens has recently been questioned. According to a new opinion - shared only by few Rubens scholars and so far totally ignored in the Leonardo literature - the drawing in the Louvre is not by Rubens but by an unknown draftsman of the sixteenth century. Rubens seems only to have enlarged, retouched and, in some places, altered the work of the sixteenth-century artist. In view of this new opinion I shall demonstrate why the Louvre drawing cannot be attributed to Rubens, why it should be dated in the first half of the sixteenth century and why it was not copied after other copies of the `Fight for the Standard'. Furthermore, I shall argue that the Louvre drawing and two of its early copies (figs. 2 and 3), which were executed before Rubens or some other draftsman did the alterations and enlargements, should be regarded as authentic reproductions after Leonardo's incomplete original version of the `Fight for the Standard'. Finally, I will discuss the main compositional features of the Louvre drawing and its early copies and demonstrate their close relationship to Leonardo's art theory and to his surviving drawings of the period between 1503 and 1506, that is, of the period when he worked on the `Battle of Anghiari'.
FROM SOURCES and documents which pertain to the `Fight for the Standard', we know that Leonardo received his commission for the painting in autumn 1503 and that he was given access to the `Sala del Papa' in the Ospedale di S. Maria Novella. Here he started to work on the cartoon between the beginning of March 1504, after the paper had been prepared and cut to size (quadratura) and the end of June, after payments for `rinpastare el cartone' (re-glueing of the cartoon) are recorded. A contract of 4 May 1504 obliged Leonardo to finish the cartoon by February 1505, but it also offered an extension of this deadline if he should decide to transfer a part of the unfinished composition from the cartoon to the wall. He began to paint between 13 March 1505, when the `Duomo' Opera was asked to deliver the wood for the `ponte' in the `Sala del Consiglio', and 30 April 1505, when payments for materials and labor concerning the wall painting are recorded. In June 1505, work was well under way and payments for the painting are documented until 31 October 1505. Leonardo stopped working on the `Battle of Anghiari' at the end of May 1506, when he received permission to go to Milan for three months, and he seems not to have resumed his work on the incomplete painting after this date. On the contrary, according to later sources such as Paolo Giovio, the Anonimo Gaddiano and Giorgio Vasari, Leonardo left the painting unfinished because of problems with a new but totally unsatisfactory technique of painting.
In the first half of the sixteenth century, the most important visual source for any copyist of Leonardo's composition of the `Fight for the Standard' were the cartoon and the wall painting. Leonardo's then-extant preliminary sketches, which showed some ideas for this composition, would not have been easily accessible between Leonardo's departure for France in 1516 and Francesco Melzi's death in 1572. Another source for a copyist of the `Battle of Anghiari' may have been the so-called `trial panel', first mentioned by the Anonymo Gaddiano, who also describes its particular purpose. According to his account, Leonardo, who wanted to test a new technique, had painted in oil on a wooden panel which he had prepared with plaster. Then he dried the colours by the heat from a fire lit in front of the panel. This `trial panel' almost certainly did not include any significant part of the composition, because Leonardo, who in those days showed an increasing reluctance to paint, would not have combined a technical test with any elaborate painted composition. It is highly dubious that the `trial panel' could be of any importance for our knowledge of the `Fight for the Standard'. Therefore I shall consider only the wall painting and the cartoon as significant visual sources for early copyists of Leonardo's `Fight for the Standard'.
I turn now to consider briefly the extant copies, which differ in various details and, most significantly, in the rendering of the standard and in the representation of the fourth rider. In fact, in most copies, the object of the battle, the top of the Milanese standard with the banner, is almost completely missing. Only its shaft, which extends across the whole composition, is clearly visible in all the copies. In the so-called `Tavola Doria' (fig. 5) and in the Uffizi-panel now in the Palazzo Vecchio (fig. 6), the shaft of the banner is unfinished on the right side and the banner itself is missing almost entirely. Also in these copies the rider to the far right appears in outline and only his head is executed in detail. In the Uffizi-panel the same rider at least shows more detail in his sinopia-like inner and outer lines. These lines indicate vestiges of the banner on his shoulder, a spear in his right hand, a slightly curved sabre to his left side, some footwear and parts of the horse's saddle. Thus, it seems as if the copyist of this panel had copied preparatory strokes of an incomplete figure (also the rather awkwardly cut-off legs of the horses in the Doria and the Uffizi panels suggest that these copies were both reproduced from an either unfinished or damaged work). The incomplete state of the banner and of the Milanese rider at the far right is also, although less obviously, recorded in an early drawing in the Rucellai Collection (fig. 7) and by the drawing in the Louvre and its duplicates (figs. 1-3). The reliability of the Rucellai and the Louvre drawings, however, is slightly compromised by some alterations: the artist of the Rucellai copy equipped the Milanese rider with dress, sword and armour, and in the Louvre drawing not only this rider's martial outfit but also the missing banner have been added. In the copies after the Louvre drawing (which I shall discuss below), there are fewer alterations of this kind and these copies consequently give a more reliable rendering of the incomplete original composition.
From these observations, I would draw the following conclusion: the Rucellai drawing, the Palazzo Vecchio panel, the `Tavola Doria' and the Louvre drawing and its copies give evidence of an original composition which was unfinished at the upper right side, where parts of the fourth rider and of the banner are left out. The representation of this unfinished condition of the original composition suggests the high reliability of this group of copies.
II The best copy and its copies
The Louvre drawing, traditionally attributed to Rubens, measures 452 by 637 mm and is executed in black chalk and pen and ink. At some later stage, the drawing was heightened with lead white, blue, grey and white body-colour and also enlarged on all four sides (the inner sheet measures c. 420 by 577 mm). Both the underdrawing in black chalk and the work in pen and ink are confined to the inner sheet. The retouchings and alterations in wash, lead white and body-colour, however, have been applied to the already enlarged sheet and in particular the extensive layers of grey and blue body-colour cover the whole drawing evenly. The most obvious additions are the right arm of the third rider, who brandishes a sabre (applied with brush and colour), and the tail of the horse to the right. These additions also occur in three further copies: 1) in a drawing in Cambridge (Mass.; fig. 12), 2) in a print after the Louvre drawing (fig. 13) by Gerard Edelinck (1640-1707) and 3) in a Rubens painting in Vienna done after the same composition. It follows that these three works were executed subsequent to the enlargements and alterations of the Louvre drawing. More important yet are another two very exact copies after the Louvre drawing in The Hague (fig. 2) and in Los Angeles (fig. 3), because these copies were done prior to the enlargements and alterations of the sheet in the Louvre. They are crucial for our knowledge of the original, unaltered appearance of the Louvre drawing, i.e. for its appearance in the Cinquecento when it was produced.
Prior to its acquisition by the Louvre in 1852, the drawing had passed through at least five different private collections, mainly in Sweden, and was therefore known to only a few writers in central and southern Europe. Authors in the eighteenth century seem to discuss only Edelinck's print (fig. 13) or the copies mentioned above (figs. 2, 3 and 12) and they considered these copies as inferior or mannerist works after Leonardo by an unknown Dutchman. The earliest recorded attribution of the Louvre drawing (or of one of its copies) to Rubens seems to derive from a statement by Carlo Amoretti who in 1804 voiced a current opinion that Rubens had executed the drawing and, in doing so, altered Leonardo's original composition. Later writers such as Giuseppe Bossi in 1810 or Stendhal in 1817 and most other authors in the nineteenth century took for granted the attribution Amoretti had based on hearsay. Thus as early as 1836, the copy after the Louvre drawing now in The Hague was also catalogued as Rubens, and some years later, in 1854/1856 and in 1878, the Louvre drawing itself was definitely ascribed to Rubens, without, however, giving any reason for the attribution. In 1892, Max Rooses expressed the belief that Rubens' authorship could be accepted on the basis of the dynamic and powerful forms of the composition. Most authors followed this attribution without venturing a precise argument. Frits Lugt in 1949 maintained that the drawing showed Rubens at his best in the parts done in pen and ink and in the retouching with body-colour. However, Lugt also observed that no other Rubens drawing from his Italian period shows a comparably high level of diligence. Julius Held stresses the `brilliancy of Rubens' style, especially in the work in ink and in water-colour', but he also considers that the underdrawing in chalk may be the work of a pupil.
FROM THE HISTORY of the Louvre drawing and from the few judgements quoted here it becomes clear that its attribution to Rubens is based only on descriptive observations and not on a thorough and critical comparison of its technique and style with authentic Rubens drawings. The need for a more critical examination is obvious from the judgements cited above. Lugt wondered about the unusually high degree of diligence and Held even advanced the idea that the chalk underdrawing could be the work of a pupil of Rubens. Yet another point should have prompted a more cautious view of the attribution: according to the traditional understanding, Rubens executed the composition on the inner sheet of the Louvre drawing during the first years of his Italian period. This early version in black chalk, as yet lacking the alterations and enlargements, was then copied either by himself or by a pupil, and only later did Rubens rework the enlarged drawing with a different technique. Since the stylistic features of the reworkings are different from both the underdrawing in black chalk and from the work in pen and ink, these reworkings must pertain to a later period of Rubens' career. Following this traditional understanding, the stylistic und technical contrasts in the Louvre drawing have to be explained with the rather complicated assumption that Rubens reworked his drawing not only twice (in pen and ink, and then with brush, body-colour and lead white) but also in two stylistically very different ways. Obviously, this somewhat difficult reasoning - an artist copies a copy of an unfinished composition by Leonardo, copies his own copy (or has it copied by a pupil), enlarges and alters the first copy and finally completes the unfinished parts of his original composition (i.e. the first copy) - suggests another explanation: the Louvre drawing is the creation of a sixteenth-century artist that Rubens reworked. If one accepts this explanation, a more plausible sequence for the different versions after the `Fight for the Standard' results: first the Louvre drawing, without its enlargements and alterations, was produced by an unknown Cinquecento draftsman after Leonardo's original composition; then at some later stage two unknown artists of the sixteenth century copied this drawing, thus creating the versions now in The Hague (fig. 2) and Los Angeles (fig. 3); at the beginning of the seventeenth century, Rubens acquired, repaired and altered the original Cinquecento drawing now in the Louvre (fig. 1); later he painted the version of the `Fight for the Standard' (now in Vienna) after the drawing which he had altered some years earlier; finally, after Rubens' death, Edelinck produced his print (fig. 13) after the Louvre drawing. At some point towards the end of this sequence an unknown artist of the seventeenth century created the drawing which is now in Cambridge (Mass.; fig. 12). Furthermore, if one accepts this explanation and this sequence of copies, the recognition of the Leonardesque quality of the Louvre drawing, hitherto credited to Rubens' artistic empathy, would be understandable: the Louvre drawing owes its Leonardesque quality to the draftsmanship of the Cinquecento artist who had access to Leonardo's original composition of the `Fight for the Standard'.
A DETAILED EXAMINATION of the Louvre drawing confirms that it is the product of at least two quite distinct draftsmen. The first artist copied Leonardo's composition very carefully onto a sheet of paper that turned out to be too small for the whole composition. Later, a second draftsman enlarged the sheet on all four sides and then reworked the enlarged version with lead white, wash and body-colour. The first artist's technique reveals the attitude of a good copyist who worked slowly, carefully and without much inspiration. For example, he applied the black chalk very evenly and without the kind of vivid accentuation typical of Rubens. In this un-Rubenesque way our draftsman did the main underdrawing in black chalk. Also some preliminary strokes in black chalk, which have been obscured by the later reworkings, are visible at the first horse's front legs and at the rear legs of the horse to the right. One could regard these traces as initial attempts to find the proper position of those legs. Both the preliminary nature of these strokes and the fact that the sheet's size was inadequate suggest that the artist did not copy from another copy but from a much larger composition such as cartoon or a wall painting. At a second stage, our copyist reinforced the very detailed chalk drawing with strong but not particularly fluent lines in pen and ink. With the same technique our draftsman established the shading in pen and ink, most noticeable on the horses' backs. Also with pen and ink the artist started, but did not always finish, redrawing details such as the fringe on the first rider's dress and the mail on his arms. The reinforcements of the outlines in pen and ink follow very accurately and almost timidly the chalk underdrawing and, more significantly, they are limited exclusively to the inner part of the sheet and do not occur on the added pieces of paper. The work in pen and ink and also the underdrawing in black chalk were without doubt executed prior to the enlargements.
The particularly close relationship between the chalk underdrawing and the reinforcements and shadings in pen and ink becomes more evident from a further analysis. With his pen the artist gives a stronger edge to the outlines (most noticeably in the shield of the warrior at the lower left side) and in some places clarifies the details. This kind of clarification can be noticed in the first rider to the left, where the mail on his arm is almost pedanticly reworked with pen and ink. In other places, as for example in the fringe of the first rider's dress or in the octopus on his shoulder, the pen work remained unfinished. Thus the artist achieved a contrast between darker areas, reworked with pen and ink, and brighter parts, done only in black chalk. This procedure is consistent with his use of pen and ink for the shading, to be found in almost every part of the drawing, but most noticeably on the left side (we know in fact from other copies that the light came from the right and that therefore the left side was darker). However, the short and thick strokes in pen and ink of the Louvre drawing are unlike Rubens, who had a very different technique for rendering shades with the pen. He often used cross-hatching for this purpose (as, for example, in his preparatory drawing for the `Duke of Lerma', fig. 14), and generally his strokes are lighter, longer, more fluent and much less pedantic than the heavy dots in the Louvre drawing. In addition, the timidly re-drawn outlines of the shield on the lower left side lack the quality of Rubens, who made these kinds of reinforcements more confidently and with a more subtle application of the ink. From these observations, we can conclude that both the chalk underdrawing and its reworkings with pen and ink were executed by the same artist or, less likely but also possible, that the parts in pen and ink were done by another draftsman who followed his predecessor in an extremely careful and almost timid way. Neither draftsman can be identified with Rubens on the basis of technique.
THE REWORKINGS in wash, body-colour and lead white are of an altogether different nature, much more vigorous and with little respect for the work done in black chalk and in pen and ink. For example, in the drapery of the first rider's cloak, in the additional modelling of the two soldiers fighting on the ground, in the hand added to one of those soldiers or in the corrections to the belly and the mane of the horse to the left, the artist rigorously drew over the older composition. These additions are done energetically and they are significantly different both from the more cautious underdrawing executed in black chalk and from the almost timid reworking in pen and ink. Therefore, the author of the alterations and additions, who most likely also did the enlargements, can hardly be identical with the artist of the black chalk drawing. However, at least the reworking in wash, lead white and body-colour seem to be by Rubens. In particular the arm given to the second rider from the left shows a rendering of the muscles almost as if the arm was composed of a bundle of little flames. This kind of touch with the brush, in fact, occurs frequently in Rubens' oeuvre.
Rubens' practice of reworking the creations of other artists often goes so far as to alter or obliterate parts of the original composition. Thus, in the Louvre drawing he eliminated the head of the first rider's horse and he also changed a significant detail of the foot-soldier falling onto his shield: the copies in The Hague and Los Angeles (figs. 2 and 3) and also Zacchia's print (fig. 10), show a smaller shield, and at the bottom of the sheet, below this shield, a cap that had fallen off the foot-soldier's head is still visible. Rubens eliminated the few and probably already-fading vestiges of this cap and misinterpreted a small part of its circumference as the lower perimeter of the shield. The cap thus vanishes and becomes part of the enlarged shield. This misinterpretation and obliteration of the cap suggests that Rubens was not the author of the chalk underdrawing and that he reworked the Cinquecento drawing without knowing other copies such as Zacchia's print.
Other evidence confirms that the original author of the Louvre drawing was not Rubens. The `Fight for the Standard' fits neatly into the large group of drawings repaired and retouched by Rubens who, particularly in his early career, had acquired and then, when it seemed convenient, repaired and altered drawings of earlier masters. Rubens' practice of repairing and retouching works of earlier masters was well known to eighteenth-century connoisseurs and it has received greater recognition in recent years. The `Fight for the Standard' in the Louvre is one of these drawings because it evidently shows all their typical features: repairs of the paper, emendations, substantial reworking and considerable stylistic discrepancies.
Finally, a further stylistic observation, namely, that the style of the Louvre drawing does not fit Rubens' very individual way of copying, speaks against the traditional attribution. In fact, an examination of authentic Rubens copies after older masters shows, in almost all instances, that his own personal style remains evident. Particularly in his copies of complete compositions and of compositions with several figures, Rubens emerges as a creative copyist who either altered the original composition or at least left recognizable traces of his personal style. For example, in his drawing `Augustus and the Sibyl' after Pordenone (whose authenticity is documented by the handwriting in the upper left; fig. 21), the light and fluid brush work of the water-colours bears witness to Rubens's hand. In his copy after Mantegna's `Triumph of Caesar', the soft touch around the eyes and around the chin as well as the heavy eyelids of the figures are characteristic of Rubens' faces. Other examples, such as the drawings after Giulio Romano's `Battle of Constantine' and after Paolo Veronese's `Feast in the House of Simon' (fig. 22), show that Rubens did not copy slavishly. In particular, faces often lose some of their original character and become at least slightly Rubenesque. Thus Rubens, who was always conscious of his own style, did not strive for absolutely accurate replicas of other artists's works in his copies and particularly not in the rendering of facial expression and in the re-creation of compositions with many figures. Only in his copies of small compositions or of single figures, such as in his drawings after Hans Holbein's `Dance of Death' or Michelangelo's frescoes in the Sistine Chapel, are the vestiges of Rubens' personal style hard to find.
TAKING INTO ACCOUNT these observations about Rubens as a copyist, it is hardly conceivable that he would have copied accurately an unfinished work such as an imperfect and probably rather poor copy of Leonardo's incomplete composition for the `Fight for the Standard', that is, without leaving some vestiges of his own style. Neither would he have copied embarrassing mistakes like the unsupported spearhead under the head of the third horse, which still can be seen in the Louvre drawing, but which is eliminated in Rubens' own painting in Vienna after this drawing. The particular (that is, faulty) treatment of the misplaced spearhead in the Louvre drawing also strongly suggests that this drawing is not a copy after earlier copies: if the author of the Louvre drawing had used, combined and copied other, earlier copies of the original composition, he would not have copied an obvious mistake. The use of more than one source should have excluded a mistake like the unsupported spearhead because a copyist who draws on several sources strives for completeness and accuracy and hence would correct embarrassing mistakes. Another mistake, the missing dagger of the soldier fighting on the ground, mentioned by Vasari and known from the Rucellai (fig. 7), Timbal (fig. 11) and Horne-copies, has not been eliminated, and therefore this mistake also suggests that the Louvre drawing was not copied from a combination of other copies (such as the ones mentioned above) and completed according to Vasari's description.
If one accepts that Rubens' copies after complete compositions of older masters generally show recognizable vestiges of his personal style, then it becomes clear that the human and equine physiognomies in the Louvre drawing can hardly be the creation of the Antwerp master. The faces of the two horsemen to the left, with their hooked noses and thick creases around the mouth, are in no way Rubenesque. They very much resemble Leonardo's well-known type for an old man (see below). In comparison, the physiognomy of the horses in the Louvre drawing is less Leonardesque but, again, it is equally distinct from Rubens's favorite type of horse, which usually has a much more slender head (see, for example, the drawing for the `Duke of Lerma'; fig. 14).
III The question of reliability
FROM THE ABOVE considerations we can conclude that the Louvre drawing was originally by a Cinquecento master and that it was enlarged and reworked with wash, body-colour and lead white by a later draftsman, most likely by Rubens. Rubens's alterations in some places obscure the original Cinquecento drawing and a layer of blue and grey wash covers the background almost entirely. Fortunately, we still have precise knowledge of the drawing's appearance prior to the alterations because it was copied very accurately at least twice before Rubens began to rework the older sheet. The two drawings in The Hague (fig. 2) and in Los Angeles (fig. 3) are extremely faithful copies; they give all details exactly and they also have roughly the same size as the Cinquecento part of the Louvre drawing, i.e. the central sheet without the pieces of paper added later by Rubens. In the following, I shall argue for the fidelity of the drawings in the Louvre, The Hague and Los Angeles to Leonardo's original conception.
The drawing in The Hague has been credited at least twice with having the quality of a Cinquecento work of art. The same is true of the drawing in Los Angeles which, when in a private collection in Paris around 1929, was attributed by its owner to Leonardo himself or at least to some sixteenth-century draftsman. At a sale in Paris in 1977, the same drawing reappeared under attribution to Giulio Romano. The source for both the version in The Hague and in Los Angeles was the Louvre drawing prior to its enlargements and reworking, because preliminary strokes in chalk around the front legs of the horse to the left and around the rear legs of the horse to the right can be found only in the Louvre version. In comparison, the drawing in The Hague has a different character: the tentative strokes are completely missing. That this work is a copy of another drawing - most likely copied from the Louvre version - also becomes apparent from the simplifications at the neck of the second horse (the shade between neck and head is transformed into a meaningless flat shape) and at the dress of the soldier at the lower left side (the bands of the dress are mistakenly fixed to the shield). The version in Los Angeles also seems to be a copy after a drawing, but without knowing the original I would not venture any further judgement.
The drawings in the Louvre, in The Hague and in Los Angeles constitute a homogenous group of copies after the `Fight for the Standard', whose archetype is the Louvre version in its unaltered condition. One important and striking feature of all three drawings, as discussed above, is their conscious reference to unfinished parts at the upper right side of Leonardo's composition. These references to an incomplete original are slightly obscured by the additions in the Louvre drawing, but they are clearly visible in the versions in The Hague and in Los Angeles. The horseman to the far right in particular suffers from being incomplete: his right hand with the shaft of a spear is hardly visible, the piece of cloth on his shoulder could be either a coat or a part of the Milanese banner, and the visible part of his leg (which is probably just a copy of the leg of the horseman to the far left) remains without a proper stirrup to rest on. The incompleteness of the rider to the right stands in a vivid contrast to the horseman on the left which shows an extraordinary amount of detail. This incompleteness of the right side of Leonardo's original composition is confirmed by the `Tavola Doria' and the panel in the Palazzo Vecchio (figs. 5 and 6).
There are, however, some inaccuracies in this group of drawings. For example, the straight sword of the horseman to the right is certainly drawn according to the phantasy of the copyist, because all other copies give a curved sabre. The second rider from the right, i.e. the Florentine horseman in the background, poses another problem because his head gear in our group of drawings is very different from the dragon helmet given in most other copies. However, Lorenzo Zacchia's print of 1558 (fig. 10) shows this figure wearing a turban, and physiognomies similar to that of the Florentine horseman are known from other Leonardo drawings. It remains therefore possible that the appearance of this rider reflects one of Leonardo's original concepts.
The remaining major inaccuracies, which cannot be related to Leonardo's ideas, occur at the upper right side of the composition, between the misplaced spearhead and the head of the right horseman. But at least the awkward presence of the misplaced spearhead could also have occurred in the original composition, an infelicity, as we can conclude from other copies, Leonardo seems to have eliminated later. It is therefore not unreasonable to ask if the archetype of our group of drawings derives from a work that was altered later, i.e. after Leonardo's unfinished cartoon. In this case, the copyist would have copied a misplaced spearhead which Leonardo emended later in the execution of the wall painting. This conclusion must remain hypothetical; on the other hand, it would explain the existence of an embarrassing error in drawings that otherwise aim at correctness and reliability.
The observations made so far suggest that the draftsman of the Louvre drawing saw Leonardo's unfinished composition and that he took into account its incomplete state. This incompleteness provoked some free additions (e.g. the straight sabre) on the right side of the composition, but generally our artist tried to render all features copied as accurately as possible. In particular, the first rider to the left shows a wealth of detail which is confirmed by other early copies, for example by the `Tavola Doria' (fig. 5) and by a drawing in the Louvre showing the single horseman alone (fig. 8). The draftsman of this sheet, whose obvious objective was a high degree of reliability, gives an extremely detailed account of the original, and his striving for accuracy goes as far as depicting the horseshoe nails under one front hoof of the horse. The comparison of this copy with our group of drawings shows that these versions are generally very reliable, and only in the rendering of the drapery and the octopus on the right shoulder of the horseman are there some minor variances. But further observations confirm that our group of drawings is sufficiently accurate in most important details such as the sword with its handle in the shape of a lion's paw, the armour, the head gear, the horse's tail and the strong veins of its rear legs. Furthermore, a major strength of the versions in the Louvre, The Hague and Los Angeles becomes apparent: the impressive rendering of this horseman's Leonardesque physiognomy, which in the single rider in the Louvre is still recognizable, albeit much weaker.
ANOTHER EARLY drawing of a single horseman after the `Fight for the Standard' exists in the British Museum. It is a rather pedantic but therefore also reliable work and shows the unfinished rider to the far right (fig. 9). This drawing confirms that the incompleteness of Leonardo's original composition and the `physiognomy' of the horse are rendered correctly in the versions in the Louvre (prior to the reworking), The Hague and Los Angeles. A significant difference between these drawings and the British Museum rider emerges: here the horse's teeth, very prominently displayed in our group of three drawings, are covered by its lip and are therefore not visible. However, both Vasari's description of these teeth, and their appearance in Zacchia's print (fig. 10) and in the Timbal-copy, (fig. 11) suggest that our group of drawings is also reliable in this instance.
A comparison between the two copies of single horsemen and the versions of the `Fight for the Standard' in the Louvre, The Hague and Los Angeles points to the high degree of reliability of these drawings. Furthermore, the few differences noted suggest that the Louvre version of the `Fight for the Standard' was not copied after the drawings of a single rider.
IV Leonardo's concept of balanced motion
THE BRITISH MUSEUM drawing of a single horseman (fig. 9) draws attention to a particular and important compositional feature of Leonardo's `Fight for the Standard'. In this drawing, the rider sits awkwardly far back on his horse and this undynamic position renders his violent movement towards the shaft of the standard incomprehensible. The horseman seems to be stuck to the horse instead of being in movement towards the standard, which is the goal of his efforts. This movement is particularly conveyed in our group of drawings, where the rider grasps the standard with his left hand, thus moving away and being pulled out of the saddle. A faint trace of his being pulled away by the force of the struggle is also visible in the Uffizi copy (fig. 6), since here as well the rider rises slightly from the saddle, pressing his leg and knee against the neck of the horse.
The highly-developed understanding of balanced movement and the coincidence of details described so far suggest that the prototype of our group of drawings (most likely the version in the Louvre, traditionally attributed to Rubens) has been copied from Leonardo's original composition. In particular, the convincing and accurate rendering of balanced movement, almost entirely absent from all other copies, confirms the highly Leonardesque character of these drawings. Significantly, the balance between the major forces centers around the motif of the struggle, i.e. around the standard which for only a short and transitory moment binds together the movements of the horsemen. This balance is transitory because the shaft of the standard seems to be breaking under the strain of opposing motions. It breaks firstly behind the back of the first horseman and secondly - at least in the Louvre version - further to the right, close to the hand of the fourth horseman. It remains, however, unclear whether the standard breaks twice, because in the drawing in The Hague there is no indication of a crack. But evidently our group of drawings shows the battle at a moment when a clearly visible balance of forces is compelled to break up in the next second.
From a further analysis of Leonardo's works, we can deduce that it was exactly this breaking up of balance he wanted to demonstrate. While working on the `Battle of Anghiari' between 1503 and 1506, Leonardo developed a renewed and strong interest in the problem of balance, even planning to write a book on this subject. Leonardo's increasing concern for problems of equilibrium in motion also emerges in a number of drawings from the period between 1503 and 1506, for example in his preparatory sketches for the `Battle of Anghiari' (figs. 16 and 17). These drawings demonstrate Leonardo's striving for a compositional balance between the opposing movements of a fierce and violent struggle. In the first sketch, the battle seems to be a cluster of fighting men and beasts, held loosely together without a proper center of balance (fig. 16). This center of balance is added in a second sketch (fig. 17), which shows almost the final solution for the `Fight for the Standard'. Two horses with their legs interlocked are placed in the center, and the counter-movements of the fighting figures and animals to each side are held together by the banner and its shaft. Precisely this particular device of merging movement and counter-movement into a transitory compositional balance can be found in a number of other drawings of the same period, for example in Leonardo's drawing of `Neptune with the Sea Horses' (fig. 18) or, less violently, in some sketches for the kneeling Leda. In particular, the Neptune drawing shows exactly the balance of movement which is about to disintegrate in the `Fight for the Standard'.
The Leonardesque character of the group of drawings in the Louvre, The Hague and Los Angeles is not manifested in their exemplifying Leonardo's concept of balanced movement alone. There are also other Leonardesque features, such as the rearing horses which are very similar to Leonardo's horses for the Trivulzio Monument (fig. 19). But above all, the fierce physiognomy of the Milanese horseman breathes the spirit of faces very typical for Leonardo. In fact, no other copy shows a similarly strong and Leonardesque type of facial expression, and a comparison of the Louvre and the Los Angeles drawings with Leonardo's preparatory study for one of those faces (fig. 20) or with the sixteenth-century fragment done after Leonardo's cartoon (fig. 15) suggests that our group of drawings must be extremely close to Leonardo's original idea in this point as well. Thus, in the most difficult and impressive parts, i.e. in the motif of balanced movement and in the human physiognomy, the drawings in the Louvre, The Hague and Los Angeles surpass all other copies in quality.
FOR MORE THAN A CENTURY the Louvre version of the `Battle of Anghiari' has been regarded as an example of the Flemish Baroque, and therefore one might find it hard to accept that this drawing gives a reliable rendering of an early Cinquecento work of art. However, looking at his drawings for the Trivulzio Monument (fig. 19), it becomes clear that at the beginning of the sixteenth century, Leonardo achieved a compositional quality which much later would be categorized as baroque. Obviously, aside from Raphael and the draftsmen of our group of drawings, most artists found this quality particularly difficult to capture (therefore we only have a few, and for all that, bad copies after the `Fight for the Standard'). But drawings with comparable monumentality and with a similar rendering of the figure in the round are known from Cinquecento artists such as the Anonymo Cantabrigensis, Baccio Bandinelli, Benvenuto Cellini, Bronzino or Giulio Romano. In fact, the latter has been associated already with the version of the `Fight for the Standard' in Los Angeles (see above). Also, some features typical for the Cinquecento or for Leonardo in particular have been noticed both in the The Hague copy and in the Louvre drawing.
Even granting the arguments developed above, it might still be difficult to see a drawing, which has become almost a symbol of the Baroque, as the work of a sixteenth-century artist. Indeed, only a chemical analysis of the paper of the drawing in the Louvre could produce irrefutable proof for my hypothesis that this drawing and its copies are sixteenth-century reproductions of Leonardo's composition. However, today our conception of Leonardo's composition is generally defined by the drawing traditionally attributed to Rubens, and in the literature of the last 100 or so years, my hypothesis has already found wide spread approval: the judgement of the eye has succeeded and most books on Leonardo illustrate the `Battle of Anghiari' with this powerful drawing. Sometimes images speak for themselves, and sometimes they need a little help.
* I wish to thank Jack Wassermann (my most ferocious critic), Martin Warnke, Matthias Winner and W. J. Wegener for the advice and help they gave me during the writing of this article.
ILLUSTRATIONS
1.`Fight for the Standard', Paris,Louvre, retouched by Rubens(?).
2.`Fight for the Standard', The Hague, Collection of H. R. H. Princess Juliana of The Netherlands.
3.`Fight for the Standard', Los Angeles, A. Hammer Collection.
4.Raphael, `Fight for the Standard', Oxford, Ashmolean Museum.
5.`Tavola Doria', Private Collection.
6.`Fight for the Standard', Florence, Uffizi (at present in the Palazzo Vecchio).
7.`Fight for the Standard', Private Collection.
8.Horseman from `Fight for the Standard', Paris, Louvre.
9.Horseman from `Fight for the Standard', London, British Museum (reproduced courtesy of the Trustees of the British Museum).
10.Lorenzo Zacchia, `Fight for the Standard', Vienna, Albertina.
11.`Timbal-copy', Private Collection.
12.Gerard Edelinck (?), `Fight for the Standard', Fogg Art Museum, Cambridge (Courtesy of the Fogg Art Museum, Harvard University, Cambridge,Massachusets, Gift of Edward Waldo Forbes).
13.Gerard Edelinck, `Fight for the Standard', London, British Museum (reproduced courtesy of the Trustees of the British Museum).
14.Rubens, `The Duke of Lerma', Paris, Louvre.
15.Head of a Horseman from `Fight for the Standard', Oxford, Ashmolean Museum.
16.Leonardo, `Battle of Anghiari', Venice, Accademia, no. 215.
17.Leonardo, `Battle of Anghiari', Venice, Accademia, no. 216.
18.Leonardo, `Neptune', Windsor, RL 12570 (by gracious permission of Her Majesty the Queen).
19.Leonardo, Trivulzio Monument, Windsor RL 12360 (by gracious permission of Her Majesty the Queen).
20.Leonardo, Trivulzio Monument, Windsor RL 12345 (by gracious permission of Her Majesty the Queen).
21.Leonardo, Head of Warrior, Budapest, Museum of Fine Arts.
22.Rubens, `Augustus and the Sybill', Paris, Louvre.
23.Rubens (?), Battle of Constantine, Paris Louvre.
24.Rubens, `Feast in the House of Simon', London, British Museum (reproduced courtesy of the Trustees of the British Museum).
NOTES