CODART, Dutch and Flemish art in museums worldwide

Toward the end of the nineteenth century, by which time research into early Netherlandish art was already considerably advanced and artists such as Jan van Eyck and Rogier van der Weyden had become familiar figures in European art history, a clearer picture of Hugo van der Goes’s artistic production was only beginning to emerge. Van der Goes was accepted into the Ghent painters’ guild as a master in 1467 but around 1476 withdrew to the Roode Klooster near Brussels to live as a lay brother. He died at the monastery in 1482/83. During his lifetime he had been in great demand and had achieved considerable renown, as his former confrere Gaspar Ofhuys reports.¹ Even Albrecht Dürer, during his journey to the Netherlands in 1520/21, admired Hugo’s works, referring to him a “grosz maister.” Although the name Hugo van der Goes was kept alive over the ensuing centuries by Karel van Manders’s mention of him in his Schilderboeck, knowledge of the painter’s work gradually disappeared. This situation only changed after the monumental Portinari Altarpiece in Florence was identified as the work of Hugo van der Goes based on a reference by Giorgio Vasari, who called its creator “Ugo d’Anversa.” This triptych was the starting point for further attributions from which an oeuvre of some fourteen works eventually crystallized, with discussion of some of the paintings continuing to this day.

In 1890, in Monforte de Lemos in northern Spain, Carl Justi discovered a hitherto completely unknown early Netherlandish painting depicting the Adoration of the Magi, and immediately recognized the hand of Hugo van der Goes.² It was standing on a side altar in the church of the Colegio de Nostra Señora de la Antigua, which had been founded as a Jesuit convent in 1593. Other connoisseurs visited Monforte to see the panel, and the attribution stuck. Because the convent, now in the hands of the order of Escuelas Pías, was urgently in need of money, the decision was taken to sell this evidently valuable work to the highest bidder. A sealed-bid auction was held in Madrid in 1910 and the painting went to the agent of the Berlin museums for one million reichsmark, an exorbitant sum that to this day remains the highest price paid by the Berlin collections for a single painting since they were founded.³ Nevertheless, the panel initially remained in Monforte because the Spanish authorities objected to the export of the important artwork. After much toing and froing, the sale was recognized as legal in September 1913 and the painting was shipped to Germany. It arrived at the Gemäldegalerie in Berlin at Christmas 1913.

The panel, still in its original frame, measures 170 x 263 cm, including frame, making it the second largest of Hugo’s works after the Portinari Altarpiece and without doubt one of the most outstanding creations of early Netherlandish painting as a whole. Although generally known as the Monforte Altarpiece, it is only the central panel of what was originally a triptych, as betrayed by the hinges that remain in place on either side of the frame. Of the wings, which likely depicted the Birth of Christ on the left and the Presentation at the Temple or Circumcision of Christ on the right, no trace remains.⁴ Indeed they had become separated from the central panel by the time of its installation in Monforte around 1600, for the painting was then housed in a gilded aedicula, with side columns and a heavy architrave, that left no room for wings.⁵ It is possible that they had not even accompanied the panel when, under unknown circumstances, it had been transported to Spain in the middle or second half of the sixteenth century. And yet, as it turned out, these wings were not the only elements the work had lost.

The panel reached Berlin as a regular, broad rectangle. At the museum, however, it was noticed, as early as the beginning of 1914, that the upper beam of the frame displayed two diagonal joints or miters that separated the approximately 77 cm wide central part from the beams to the left and the right. The central portion was then removed to reveal a 10 cm high painted strip. This was the remnants of a top extension that would have given the panel an inverted T-shape, as was common in large Netherlandish retables of the fifteenth and early sixteenth centuries. A reason for the removal of the central extension can be inferred from the architectural framing of the panel in Monforte. At its original height, the panel would have fitted into the aedicula made around 1600. In 1666, however, a second, smaller aedicula was made in which a painting of the Jesuit saint Francisco de Borgia was presented. This was positioned on top of the existing aedicula.⁶ In order that one extension should not, as it were, sit above the entablature of another (that of Hugo’s painting), the extension of the Netherlandish panel was sawn off flush with the upper edge of the frame, and the horizontal beam of the frame was inserted into the resulting gap below.

In Berlin it was decided to make the newly discovered remnant of the extension visible, and to this end two short (merely 10 cm high) frame sections were made and joined to the surviving original upper beam of the extension. This unit was then placed on top of the panel, effectively giving it a central extension measuring 77 cm wide by a mere 10 cm high. This central extension is not particularly successful in terms of its proportions but it does at least, thanks to the garments of the angels, allow important parts of the lost composition to be imagined. This is the form in which the Monforte Altarpiece became known and in which it was on view at the Gemäldegalerie in Berlin between 1914 and 2022 (fig. 1).

What was once to have been seen on the detached part of the extension, which has disappeared without trace, was ascertained in the very first publications on the newly discovered panel.⁷ The specific motifs could be read beyond all doubt from a number of free copies and variants of the composition. A large drawing by Hans Holbein the Elder and two paintings from the workshop of the Master of Frankfurt, who was active in Antwerp shortly after 1500, are unanimous in showing the upper part of the ruins with a broken arch, a rear tower with corner turrets and a dilapidated roof truss, as well as two groups of small, hovering angels.⁸ Nevertheless, the first visual reconstruction of the extension was not attempted until 2002, when Rainald Grosshans combined a black and white photograph of the surviving panel with a drawn central extension.⁹

During the course of preparations for the first Hugo van der Goes exhibition, which opened at the Gemäldegalerie on 29 March 2023, I had further thoughts about the painting’s lost extension. My starting point was the idea that the proportions of the entire work, and thus its overall effect, would alter considerably with such an extension in place. Using Photoshop, I therefore made a reconstruction that was intended to be, and indeed was, reproduced in the catalog (fig. 2). I was able to take the motifs from the free copies of the panel, as Grosshans had previously done. I also took my cue from Grosshans with regard to the height. As was often the case in retables of an inverted T-shape, the extension needed to be square. Moreover, knowledge, unavailable to Grosshans, of the aedicula made in Monforte in around 1600, also came to my aid. Its internal dimensions allowed exactly enough space for an extension of 77 cm.

The effect on the original painting of the extension created in Photoshop was astonishing. Whereas Hugo’s truncated panel, despite its artistic perfection, seemed curiously cramped and confined, it now became much airier, seeming to allow the enormous figures the space they needed. This led to the idea of attempting at least a temporary reconstruction that would remain in place for the duration of the exhibition. The original plan was to scale up the Photoshop collage to the dimensions of the lost original and position it above the panel while intending that the reconstructed part should be easy to identify. The extension would be bordered by two new frame strips, exact copies of the originals. These frame strips were perfectly made and gilded by Bertram Lorenz, the Gemäldegalerie’s restorer. In place of the collage, however, Lorenz then suggested creating a painted reconstruction of the extension, which would match the original more closely. Before he started work, we agreed that the painted reconstruction would not be used unless we were entirely satisfied with it. Once Lorenz had begun, I was still very much thinking that even the painted reconstruction should be differentiated in some way, for example by means of hatching or by covering it with a diffusive film.

Bertram Lorenz then painted a reconstruction in oils, as far as possible taking his cue from Hugo’s original. He sensitively developed the pattern of folds in the garments of the two foremost angels from the surviving lower portions of their draperies, and for the faces of these two angels he took as a model the heads of the angels in the Portinari Altarpiece. With regard to color, the reconstruction needed to precisely match the tones of the original painting, which proved particularly difficult in the case of the sky and the yellow garment of the foremost angel on the right. Lorenz was able to execute his reconstruction in front of the original but only occasionally could he place it above the panel for comparison purposes (fig. 3).

The results of his work convinced all those involved in the project: in terms of color and structure, the sky, architecture, and garments integrate seamlessly with the original painting. The individual motifs in the new extension must surely at least come close to what was originally there, and harmonize fully with the overall composition of the work. In the light of the highly successful reconstruction, we also decided against differentiating the reconstructed part by hatching or other means because this would have detracted from the effect, especially as the color and luminosity of the reconstruction would have changed. The Monforte Altarpiece, with reconstructed extension, was eventually installed in the exhibition space at the end of March 2023 (fig. 4).

Naturally, even before the reconstruction there had been discussion within staff circles of the legitimacy of complementing a 550-year-old original in this way. From a purist standpoint, it could be seen as falsifying a historical state, and a few decades ago the project would probably have been rejected. As is often stated in restoration ethics, the possibility to restore a work to its original condition is an illusion. But this was not the aim of the reconstruction, particularly as significant parts of this work—the wings—had already been lost. Rather, it was to allow the composition of the central panel to unfold in the manner originally intended. The small patch of sky and the summits of the architecture above the centre of this Adoration of the Magi make the entire picture space seem higher and more open. The previously low, almost basement-like ruins now seem tall and extensive, although they are in fact very small in relation to the figures.

In the exhibition, the modern reconstruction is identified as such in text panels and a diagram, and the join between the original panel and the addition is clearly visible. The wall texts state that the reconstruction will only remain for the length of the exhibition. Since the exhibition opened, however, numerous colleagues and other visitors have pleaded for it to be retained even once the painting is again displayed as part of the collection. Accordingly, we have decided to leave the reconstruction in place for the time being. It has been installed without any intervention in the original substance and is therefore fully reversible. What remains to be seen is how long it will last.

Dr. Stephan Kemperdick is Curator of Early Netherlandish and Early German Painting at the Gemäldegalerie, Staatliche Museen zu Berlin in Berlin. He has been a member of CODART since 2005.

This article was translated from German by Richard George Elliott.

¹ On the history of research into Hugo van der Goes, cf. Stephan Kemperdick, “Hugo van der Goes: The Loss and Rediscovery of an Extraordinary Artist,” in Stephan Kemperdick and Erik Eising (eds.), Hugo van der Goes. Between Pain and Bliss, Berlin, Gemäldegalerie 2023, pp. 11–17. For Ofhuys’s text ibid., pp. 276–279.
² See Max J. Friedländer, “Die Anbetung der Könige Hugos van der Goes,” in Jahrbuch der Königlich Preussischen Kunstsammlungen 35 (1914), pp. 1–4; later Antonio Mendez Casal, “Les gandes œuvres d’art inédites,” in Les Arts anciens de Flandre, vol. IV, Bruges 1909/10, pp. 156–161.
³ Erwin Hartmann, “Bilder haben ihre Schicksale. Ankauf der ‘Anbetung der Könige’ des Hugo van der Goes vor achtzig Jahren,” in Jahrbuch Preußischer Kulturbesitz 30 (1993), pp. 503–514.
⁴ Cf. Hugo van der Goes. Between Pain and Bliss 2023, cat. 2, 4.
⁵ Cf. Ibid., cat. 2.
⁶ Ibid., p. 119, fig. 3.
⁷ Joseph Destrée, Hugo van der Goes, Brussels 1914, pp. 77–89, 211–213, 233–237; Max J. Friedländer, “Die Anbetung der Könige Hugos van der Goes,” in Jahrbuch der Königlich Preußischen Kunstsammlungen 35 (1914), pp. 1–4.
⁸ Cf. Hugo van der Goes. Between Pain and Bliss 2023, cat. 2, 3, 4.
⁹ Rainald Grosshans, “Hugo van der Goes (um 1440–1482). Die Königsanbetung aus Monforte in der Berliner Gemäldegalerie,” in Jahrbuch Preußischer Kulturbesitz 34, (2002), pp. 131–164.