Rogier Van Der Weyden's Philadelphia 'Crucifixion'

Home , Crucifixion, Rogier van der Weyden, Triptych

MARK TUCKER

Rogier van der Weyden's Philadelphia'Crucifixion'

and founded DESPITERogier van der Weyden's stature among his con- and state, and in 1992-93 cleaning restoration, undertaken. temporaries and his sustained influence over Northern Euro- on this research, were has been almost pean art both during his lifetime and after his death in 1464, Though the painting unanimously G. much about his life and art remains frustratingly vague accepted as by Rogier since its acquisition by John John- in as to its because of meagre surviving documentation and the difficul- son 1906 (Fig.13), opinion dating, original appear- has been varied." A first consensus ties of confirming core works among the output of his shop ance and overall quality around the of Max who and many followers. One area of inquiry that may yet con- developed opinions J. Friedlander, and treatment are tribute much to a sharper definition of his artistic identity noted in 1921 that the subject sculptural but then stated that the two is the study of his materials and technique, especially as 'entirely of Rogier's genius', pan- is had 'formed the outer to an the information gained from the most important surviving els on which it painted wings and are in almost in source - the paintings themselves - may turn out to support altarpiece painted very light tone, grey Three in Die or to contradict the assumptions and opinions that still grey except for the red curtains'.4 years later, added its concen- remain the basis for most judgments concerning works in the AltniederliindischeMalerei he only: 'In simple a characteristic invention of the Van der Weyden group. Such study has significantly trated design highly Master.'5 This and the of the colour as improved the understanding of one work in particular, the judgment description nearly gri- combined with as to whether Crucifixionwith the Virginand St John in theJohn G.Johnson Col- saille, scholarly disagreement format the lection at the Philadelphia Museum of Art (Fig.23), which has the large bipartite represented altar-piece wings, a or even of an elicited varied and sometimes directly conflicting scholarly left and centre sections of triptych parts organ the that the two were opinions through the decades. Research carried out between case, gave impression panels subsidiary 1981 and 1990 has done much to amend and consolidate our elements from an incomplete work.6 The acceptance of such contributed to a relative of the understanding of this work, settling some questions that estimations perhaps neglect over the decades. could previously be addressed only speculatively. panels following view of the and its Technical investigation of the painting began in 1981 with An alternative painting's quality impor- of and infra-red reflectography conducted for a survey of under- tance to the understanding Rogier's style chronology in 1939 with Richardson's that the drawing in paintings by the Master of Flkmalle and Rogier emerged E.P. proposal is the 'Cambrai a work for which van der Weyden. The survey, published in 1992, singled out Johnson painting altar',7 is some He noted that the Philadelphia Crucifixionas a pivotal painting within the there contemporary documentation.8 the work's format Rogier van der Weyden group, in that its underdrawing com- nearly square corresponds approximately to the dimensions in the and that its bines key features of the distinctly different underdrawing given documents, argued size and make it one of the styles of two benchmark works, the Prado Descentfrom the Cross 'great superb quality clearly major efforts of with the artist's record- and the Escorial Crucifixion.'Though the study's investigation Rogier's career', according involvement in the of the to Cambrai. of materials and technique was limited, it did produce some ed delivery altar-piece for the and observations, both technical and qualitative, indicating Rogi- Richardson's argument painting's importance in er's direct involvement in the actual painting of the Johnson quality depended part upon overturning misconceptions about its and he observed that con- Crucifixion,while dendrochronological analysis of the panels colour, pointedly 'quite to of it scholars as almost a supports the late dating assigned to the work by a number of trary descriptions by European scholars since the 1940s.2 Further examination of technical, monochrome painting, it is of a special brilliance of color'. Documents date the Cambrai altar to the between documentary and stylistic evidence, carried out in 1990 in period 1455 and 1459. faced with the then the Conservation Department of the Philadelphia Museum June June Richardson, view that the was a of Art, yielded new information on the painting's technique prevailing scholarly Johnson painting

to with for the J.R.J. VANASPEREN DE BOER, J. DIJKSTRAand R. VANSCHOUTE: Underdrawing in Paintings tiously positive strongly enthusiastic, Roger Fry, example, praising as 'one of the most of theRogier van der Weyden and Master of Flmalle Groups(Nederlands Kunsthistorisch Jaar- panels in his 1906 correspondence with Johnson impressive in a letter of boek,XLI), Zwolle [1992] pp.25, 37, 158. creations of Medieval thought'. MaxJ. Friedlander, toJohnson 1907, called them that he had tried to them for the Kaiser 2Ibid., pp.156 and 158. In the 1981 comparison of the underdrawing, the few older 'magnificent,' adding acquire His later was more reserved. X-radiographs then available and the paint surface, two creative moments were Friedrich Museum. published opinion, though positive, of van der in Art in Amer- noted - in the underdrawing and in a first paint layer - suggesting Rogier's direct 4M.J. FRIEDLANDER:'The Pictures Rogier Weyden America', involvement in at least those stages of execution. Further examination of new full sets ica,IX [1921], p.65. van der und der Meister von of X-radiographs and the painting itself in August 1992, however, led van Asperen 5Idem:Die AltniederliindischeMalerei, II, 'Rogier Weyden de Boer to observe that 'there are corrections even in the top paint layer', and to con- Flkmalle', Berlin [1924], p.96, no.15. MA note 2. The clude: All this [the continuous development of the figures throughout the drawing, 6E. PANOFSKY:Early NetherlandishPainting, Cambridge [1953], p.285, of the format was established Col- underpainting and final paint layers] points to Rogier's own hand in all stages and completeness bipartite eventually by theJohnson Curator Henri Marceau's observation of the of the probably no participation of the workshop at all'. (Correspondence of 25th August lection symmetrical positioning the division between the two than to the outer 1992 in the painting's conservation file.) He also noted the presence of a thin, cloths of honour closer to panels translucent tan or 'flesh-colored' paint-layer applied over the underdrawing before edges. van der Cambrai Art painting was begun, a feature he had also found in the one other work studied for 7E.P. RICHARDSON:'Rogier Weyden's Altar,' Quarterly,II [1939], which specific layer structure information was available, the Prado Descentfrom the pp.57-66. St Cross.The authorwishes to thankDr van Asperende Boer for generouslysharing his 8The documents pertaining to the commissioning of a painting for the Abbey of Cambrai are in Archives du du no. knowledge of Rogier's technique and materials. Aubert at Lille, Departement Nord, 36, H431, 3Early critical responses afterJohnson's acquisition of the panels ranged from cau- fol.221. 676 Burlington Magazine is collaborating with JSTOR to digitize, preserve, and extend access to The Burlington Magazine ® www.jstor.org ROGIER VAN DER WEYDEN'S PHILADELPHIA 'CRUCIFIXION'

----i

1Y,

ii - I-1811:-'-

1A iA?,w

;? j:,?c' :i-_-_

$a-?r:~-: :? L -:--~

.. . .

13. Crucifixionwith the Virgin and St John, by Rogiervan der Weyden,as it appearedc. 1906 afterits acquisitionbyJohn G.Johnson.Oil on oak panel, 178 by 90.2 cm. (eachpanel). (PhiladelphiaMuseum of Art). work of the 1440s if not earlier,'proposed revising the stylis- Henri Marceau,then curatorof theJohnson Collection,is tic standardsfor the artist'schronology, pointing to qualities cited in Richardson'sarticle and is known to have discussed which he ascribed to the influence of Italian art following its revisionistanalysis with scholarssuch as Leo van Puyvelde Rogier's 1450 journey to Italy,such as the painting's'extra- and ErwinPanofsky. One consequenceof Marceau'sinterest ordinary monumental severity', and especially its colour in what he had come, via Richardson,to understandas the which, he observed,achieved 'an almost High Renaissance work'sessential character was the restorationcarried out in largeness and power' through 'contrastsof broad, simple 1941 under his direction by the freelance restorer David color areas'. Although Richardson's identification of the Rosen, a treatmentthat seems in some ways to have been a Philadelphiapanels with the Cambrai altar was later chal- direct, quite literalresponse to the language of Richardson's lenged,'0his stylistic observationshad significantrepercus- descriptions. This intervention changed the painting sions. markedly- more than any otherintervention in the preceding

9PANOFSKY, loc. cit. at note 6 above. the Philadelphiapanels could not have been the Cambrai or any other shutters '0PANOFSKY (loc. cit. at note 6 above, note 3) pointed out an errorin the transcription because of 'their admirablycareful execution, the sophisticatedcolouring and the used by Richardson,'un tabliau a II huystoires'implying a picturein two scenes. The presence of gold;in all otherinstances, the exteriorwings of Rogier'saltar-pieces are document actuallyreads 'un tabliau a IIhuysseries',translated by Panofskyas 'a retable treatedas second-classwork and were normallyentrusted to assistants'(ibid., note 2). with two shutters'.Though agreeingwith Richardson'slate dating,he reasonedthat 677 ROGIER VAN DER WEYDEN'S PHILADELPHIA 'CRUCIFIXION5

480-odd years- determininghow it would be known for the next five decades. No photographswere made during the 1941 treatment and the writtenrecord consists of a few brieflines in the cura- torial dossier and an article written by Rosen that year." These state that he had found the panels to be 'heavily .. streakedwith old varnish and repaints'which the cleaning undertook to remove. As for the Rosen backgrounds, 14. Detail of the Virginand St John 15. Cross-sectionof a paint sample of declaredthat the darkblue areas above the wall were '18th panel, showinga survivingpatch, the pre-1941 blue sky,showing azurite centuryrepaints' in 'an oily pigment foreignto the epoch of above the coping stone of the wall at over a darkgrey underpaintingand the that traces of describedas 'old' or the right edge, of the blue background chalkground. picture',adding gold, paint largelyremoved from the panels 'original',were discoveredbeneath. Consequentlythe blue in 1941. backgroundwas removedand the area above the wall gilded. s ~i;g;g~ ,? ;?w;?x The very brevityof the first-handaccounts of the treatment Sr : .C~B '~~~:C* lent them a forceful authority,implying that actions taken C~?- ~ "" " '~*r gY, ??P 19:a ;~ had been the singlepossible response to plain and confirmed -:k facts of the and it is therefore ;' painting'stechnique condition; :i?? not that the were without 1?r; surprising changes accepted ques- ~s~~;? S tion. Leo van 1941 discussionof the after , ? o?*ib~k~,-;b, ,I Puyvelde's painting .;?,? r ?j~*~_ treatmentshows an acceptanceof Richardson'sobservations Ik? "p;~ ~n~i?~? ;P scholarshave an idea that it is in 16. Dispersedsample of the azurite 17. Dispersedsample of azurite(here ('Most European painted found in sites throughoutthe back- with lead white)from a shadowin the grisaille,but it is in color'),and he goes on to note that 'since ground on both panels. Virgin'srobe, shown at the same magni- the work has just been successfullycleaned throughthe dili- ficationby polarisedlight as Fig.16. gent effortsof Mr. Henri Marceau and Mr. David Rosen, a backgroundof gold at the top gives an extraordinarylumi- when the settingexhibited a lack of Netherlandishdetail and nosityto this diptych'.'2With thispraise, the gold background surface refinementthat seemed to representthe surpassing and other featuresof the painting'sappearance after restora- instanceof Rogier'sreduction of a subjectto its spiritualand tion began to be consideredas absolutefacts of originalcon- dramaticcore. This apparentlyintentional abandonment of ception. the highly specificrepresentation of outwardmaterial quali- The replacementof the darkbackground with one of gold ties of the subjectneatly epitomisedPanofsky's description of did, afterall, align the paintingdirectly with othersthat have Rogier's world as 'at once physicallybarer and spiritually gilded backgrounds,notably the Prado Descentfrom the Cross richer than van Eyck's'.'5 and the Beaune LastJudgment. The gold on the Philadelphia Comparison of photographstaken soon after the panels panels provideda textbookexample of a gothic featureset as enteredJohnson'scollection (Fig. 13)16 with one takenafter the a foil to the painting'sotherwise progressive qualities, sup- 1941 treatment(Figs. 18 and 19) shows the generallyheight- porting the prevalentview of Rogier'sart as a reconciliation ened austerityof the setting and the change in effect due to of conservativesensibilities with his own powerfullyinventive the gilding of the sky.The alterations,extreme as they were, personalvision. The gildedbackground, thus seen as a fitting were never seen as violating the stylisticparadigm for the archaism,was also regarded as recovered evidence of the artist; on the contrary, the post-treatment appearance painting'simportance. The use of gold, along with colouring seemed to extend from and then support a generallymore that, as Panofskylater observed,exemplified a 'sensibilityto favourable placement of the painting within the canon. color ... second to none',"' bolstered arguments for the two Though the alterationswere accepted without question and panels'status as an independentwork, rather than altar-piece were even interpretedas components of style in the years wings. Finally,the gilded background,accepted as appropri- after 1941, the lack of any documentation apart from the ate in itself, replaced a true stylistic anomaly: the unmodulat- curator'sand restorer'ssay-so left questionsabout the factual ed dark sky unknown in any comparable work by the artist. justificationfor the restoration.Mounting concern prompted Post-War scholars attached significance and merit to the technicalinvestigation of the paintingin 1990, which also another aspect of the painting's appearance after cleaning sought to determine the possible benefits of a new cleaning that seemed to reinforce established views of Rogier's work and restorationin conformitywith present standardsof sub- and the Philadelphia painting in particular. The Crucifixion stantiation,documentation and care in execution. was already being described as 'abstract' by 1923,14 and has At the startof the 1990 examinationthe state of the back- been admired by scholars and artists alike in this century for ground was of greatest immediate interest. Lacking any seemingly anachronistic formal qualities such as the domi- record of the location or appearance of the traces of gold nance of design over detail and the broad massing of colour. describedas havingbeen found beneath the dark-blueback- The painting's severity and breadth of handling, if remark- ground while it was being removed in 1941, the search able before the cleaning, were much more so afterwards, for early gold or remnantsof the blue repaintsmentioned in

"D.ROSEN: 'Preservation vs. Restoration',Magazine of Art, XXXIV, no. 9 [1941], space'. pp.458-71. 15PANOFSKY, op. cit.at note 6 above, p.249. 12L.VAN PUYVELDE: 'The Taste ofJohn G.Johnson, Collector',Art News, XL, no. 14 16Allpre-1941 published images of the panels derive from two large-formatglass- [1st November 1941], p.36. plate negatives of very good quality,well exposed and sharp, made soon after the 13PANOFSKY, op. cit. at note 6 above, p.285, note 2. panels enteredJohnson'scollection in 1906. In 1994 an earlycolour postcard, prob- '4W. BURGER:Roger van der Weyden, Leipzig [1923], p.31: 'The treatment [of the sub- ably producedin the 1930s, was found, showingthe left panel beforecleaning. ject], abstractin itself, is increased and overchargedthrough the unrealityof the 678 ROGIER VAN DER WEYDEN'S PHILADELPHIA cCRUCIFIXION5

~? ;;; i- "~; s ,x ??- ,? " D

n i;-

a 'f .9

~xs, ~ $

X-x i ::::-~ i ~ ~a:::-FI~~3i--, 18_:i---~ -:-::$: ~I .s:~d a: g '':' ~s:::x?~??:;~ ? ~I-_ 1-n -~::~: :::::,=::,; ;~ ?~ I~

u~ .i ~wa- ~: ::I c

; i_

?= LL ~i7:? :~::::~

18. The Virginand St John panel with alterationsbrought about by the 1941 cleaning and restoration. 19. The Crucifixionpanel with alterationsbrought about by the 1941 cleaning and restoration. the restorer'saccount was undertakenby careful systematic It appeared, then, that any gold encountered in the 1941 scanningof the area under magnification.Rosen had abrad- cleaning either had not been preservedor was exceedingly ed his new gildingto reduceits brilliance,creating a continu- exiguous,perhaps no more than the sort of incidentalscraps ous networkof gaps throughwhich it could be seen that his left behind by the gildingof a frame. removalof the darkblue paint had also taken away much of While the fact that no earlier gold could be detected in the chalkground throughout the background,in some places 1990 did not in itselfestablish the appearanceof the original exposing the surface of the oak panels. Narrow margins of background, the discovery of scattered remnants of back- intact ground on which any early backgroundmight have ground paint that had survived Rosen's cleaning settled survivedwere found along the frame barbe and along the the question beyond doubt. These remnantswere the only edges of the Cross, Christ'shead and arms and across the pre-1941 materialsfound whereverthe full thicknessof the tops of the walls,but no gold other than that appliedin 1941 ground survived(Fig. 14). Ranging from bare traces to thick was found in these or the few other scatteredsites of intact patches severalmillimetres across, all exhibited a consistent ground.All paint adjacentto the backgroundwas also exam- stratigraphyof chalk ground coveredby a deep grey under- ined closelyin the expectationthat backgroundgold, typical- painting,followed by a thicklayer of coarselyground azurite ly applied before painting, should extend at least a small (Fig.15). Samples of this blue - the only blue found in the distance beneath painted edges, but no gold was found to background- were compared with samples of unquestion- have been appliedbefore the foregroundforms were painted. ably original blue from the Virgin's garment: both were 679 ROGIER VAN DER WEYDEN'S PHILADELPHIA 'CRUCIFIXION'

found to be azurite of identical particle size distribution (Fig.16 and 17). The justification offered in 1941 for the removal of the painted backgroundswas contradicted on each point by the 1990 findings;contrary to what had been believed since 1941, Rogier van der Weyden'soriginal sky had not been gold, but the deep blackish-bluethat had been mistakenfor repaintand removed.17 Rosen'sdescription of the backgroundblue as 'oily' and of the eighteenthcentury appears to have been due to his observation of prominent drying cracks in the paint. This fault, most commonly associated with unsound techniques that arose in the eighteenth century,does, however,occur in fifteenth-centurypaintings.18 The presence of such a defect, incompatiblewith a popularbelief in the perfectionof Flem- ish technique, would have lent credibilityto the restorer's declarationthat the backgroundpaint was not of the period. 20. Detail of the area around the base of the Cross after the removal of the 1941 His 'technical' carried when link- restorations,showing areas of a light tan-colouredimprimatura exposed in 1941. opinion particularweight ed with the failureof the unmodulateddark-blue sky to meet expectationsof stylisticconsistency conceived within the lim- its of 1930s scholarship.19 The _27 dark background,which is now establishedto have been both exceptionaland authentic,may be regardedas a furthermanifestation of the reach of the artist'sstylistic con- cerns and influences.An observationof particularinterest in this regardwas made in 1981 by PennyHowellJolly, who proposed a direct relationshipbetween the form and colouring ...... of the Philadelphiaand Escorial Crucifixionsand the frescoes painted by Fra Angelico and his shop for the cells at in To RAT, S. Marco Florence.20 the similaritiesJollynoted between "M...... the PhiladelphiaCrucifixion and the Mockingof Christin cell 7 at S. Marco (Fig.22)can now be added the dark sky above the wall in both worksand in other frescoesat S. Marco, making the stylistic connexion even more strikingthan when first 21. Detail of the head of StJohn after the 1993 restoration. proposed. The backgroundof the Philadelphiapainting, so readily accepted in its post-1941 gilded state as a vestige of medievalism,may have representedquite the opposite in its originalform: it can be seen as evidence of a remarkable late-career openness to new sources and advanced form, specificallythe artist'sdetermination to apply distinctchar- acteristicsof the Netherlandishidiom and his own style to Italian motifs and composition. Thus the reconciliationof traditionsand styleseffected in the Philadelphiapainting was less between gothic and mid-fifteenth-centuryNetherlandish

"Rosenand Marceauwere outspokenproponents of technicalstudy and documentary researchin questionsof authenticity.In his essay 'Conservationand Technical Research' (PhiladelphiaMuseum Bulletin, XXV, no.184 [January 1940]) Marceau stressesthe need to 'eliminatechoices predicatedsolely upon opinion and conjec- ture', and notes that the gatheringand expert interpretationof factualdata 'greatly increase the factor of safety in restorationand conservation'.Rosen espoused an uncomplicatedscientism and saw technical study as a distinguishingfeature of the 1~ emerging professionalismof the field. His treatments,however, indicate a limited aptitudeand affinityfor the disciplinedapplication of science to the study of paintings and his effortsto alignhimself with progressivepractice of the day dependedless on demonstrationsof technicalknowledge than on the intensityof his moral outrage at the perceivedmistakes of past generationsof restorers(see his articlecited at note 11 above).This, combined with an inadequatefamiliarity with technical and stylistic characteristicsof variousschools of painting,led him to purge paintingsof actual or suspectedrestoration with a single-mindednessthat often resultedin removal of originalpaint. 'Among the many occurrencesof this sort of crackingin fifteenth-centuryworks, a particularlygermane example is the Virgin'sblue mantle in the Boston Museum of Fine Art'sversion of Rogier'sSt Lukepainting the Virgin. '9Thoughthe darksky is unusual,no scholarwriting before the 1941 cleaningseems to have found it problematicenough to questionin print. 20P.H. JOLLY:'Rogier van der Weyden'sEscorial and Philadelphia Crucfixionsand their relation to Fra Angelico at San Marco,' OudHolland, XCV, no.3 [1981], 22. TheMocking of Christ,by the circle of FraAngelico. Fresco.(S. Marco, Florence). pp. 113-26. 680 ROGIER VAN DER WEYDEN'S PHILADELPHIA 'CRUCIFIXION'

than between the latter and a branch of roughly contempo- surface to features clearly present in the early photographs raryFlorentine painting that, if distinctfrom Northernpaint- and in other paintingsby Rogier showed how principalpic- ing in forms, sharedthe subjectiveessence of Rogier'sart.2' torial elements had formerlybeen balanced and enrichedby Compellingas the case may be for formal derivationsfrom refinementsand naturalisticdetail not recognisedas authen- Angelico, in the context of Rogier'ssubject - which is not just tic in 1941 nor known to be missingfor fiftyyears thereafter. the Crucifixion, but the supreme transitional moment of Rogier'sevocation of the observedworld in specificterms Christ'sdeath - the darkbackground assumes specific narra- was restrainedin the PhiladelphiaCrucfixion, and by that very tive and iconographicpurpose. The featurelessdeep blue sky fact each detail reflectedrigorous choice. What we can now signifies the darkness falling between the sixth and ninth interpretas streaksof glazes suggestingwater stainingdown hours of the Crucifixion,with the very instant of death indi- the wall were mistakenby Rosen in 1941 for darkenedvar- cated by the just-closingeyes and ashen flesh of Christand St nish and largely removed, also disruptingfine gradationsof John's staggeringposture as he moves forwardto accept the shadows cast on the wall by the cloths of honour. Patchesof weight of the Virgin,who swoonsin final collapse,her fingers green moss and rings of ochre-coloredlichens were almost interlacednot neatlyin a conventionalgesture of supplication completelyremoved, along with other touches appliedto dif- but in the fumbling grasp of a human grief that in the ferentiateindividual stones in the wall. The cloths of honour, moment eclipses piety.22The dark sky, which must have (describedin 1913, 1936 and 1939 as 'scarlet',25but post- seemed a deeply dramatic and unsettling pall to Northern cleaning as 'flaming vermilion'),2 particularlythat of the eyes then accustomedto rich architecturalor landscape set- right panel, showed the loss of some hatched and stippled tings, also suppressedthe silhouette and symbolism of the crimson glazing in the creases and shadows,with the some- Cross, emphasising Christ's corporeal suffering.The effect what greyedoriginal surface of the vermilionbroken through registeredso powerfullythat in 1936 one writer admiringly in patches.Across the bottom, notjust glazes,but largepatch- described the 'blank black background,no continuation of es of the once-continuous, smoothly blended yellowish to the cross,'and the astonishingeffect of Christ'shands 'nailed emerald opaque greens of the grassy foregroundhad been to dark,formless space'.23 removed (Figs.13, 18 and 19). This damage, caused by the The 1990 examination revealed other, less immediately pursuit of brighter colour beneath glazes that had turned apparent, changes made in 1941, which unaccountably brown and were mistaken for darkenedvarnish, randomly involvedcharacteristically Rogierian features. The singularly exposed areasof a tan-colouredimprimatura (Fig.20) that has affecting abstract power of the Philadelphia Crucfixionis at times been misinterpretedas representingsoil, as distinct based on the synthesis of an Italian compositionalbreadth from the authenticrocks extending into the paintingfrom the and Rogier'sparticular genius for the expressivedesign and bottom edge. The translucent imprimatura,applied after dispositionof compositionalelements. Although losses inter- completion of the underdrawingbut before painting began, ruptingform due to the 1941 cleaninghad long been evident, controlledthe absorbencyof the ground,but it also servedto before the 1990 examination the very strengthof the paint- reduce the contrast of an underdrawingthat in places pro- ing's design made it impossibleto imagine the importanceof vides only a generalor inchoate idea of formswhich were left a level of illusionisticcontent which had been altered or lost to be developed and refined during painting.In the shaded in the cleaning.The paintinghad alwaysrepresented an aus- side of the hillockat right, and aroundthe shadowcast by the tere extreme of the artist'sstyle, and its strikingbreadth had skullat left, the underdrawingcan be seen much as it would been commented upon before the 1941 cleaning, but the have appearedto the artistas paintingbegan. effect of starknesswas so marked afterwardthat one writer The figures escaped any significantdamage. They have observed: 'Nothing remains of the picturesque details of sufferedonly minor incidentallosses, mostly along old repairs everydaylife', and that the scene is 'so stronglysaturated with of splitsin the panels,and lightwear in some shadows,but the light that, despitethe softnessof the flesh tintsand the hues of character of their paint surfaces had been little altered by the costumes, one gets the impressionof an almost abstract cleanings (Fig.21). Panofsky attributed differences he saw patterning of pure colors'.24The 1990 examination estab- between the setting and the figures after cleaning to artistic lished, however, that the painting was originally more varied choice. His description is of a painting whose setting is chromatically and more tactile in conception than the gener- pushed in its harshness almost unbelievably from the style of al features dominant after 1941 suggested, and had original- mid-fifteenth-century Flanders: 'Our glance is blocked by a ly exhibited a carefully measured balance between spare, grim stone wall hung surprisingly with two cloths of honour imposing design and fidelity to Netherlandish representa- of flaming vermilion..,. this wall rising behind a barren strip tional specificity and rich substantiality of surface. The cor- of land.. .'. He then contrasts the stridency of'this severe yet relation of surviving fragments and traces of the original tremendously colourful background' with the figures, where

2The other frequently cited evidence of Rogier's exposure to Tuscan painting is the WashingtonPost [13th December 1936]. This piece, by the Art Editor of the Wash- affinity between his Uffizi Lamentationand the predella painting in Munich by Fra ington Post, is valuable as the lengthiest descriptive treatment of the Crucixion before of the same Angelico subject from the S. Marco altar-piece. the 1941 cleaning. In preparing the piece Skidelsky wrote to Henri Marceau similar 22A gesture, suggesting both prayer and hand-wringing, is seen in the Mag- requesting bibliographical information (correspondence in the Johnson Collection dalen the Prado the Cross.The two in Descentfrom paintings share a strong sense of an archives) and the article, written by a non-specialist for a popular audience, repre- arrested an transitory moment, largely effect of the figures' unstable postures; and it sents an interesting attempt to place the painting and its specific pictorial means in is worth noting that for the pose of St John in the Philadelphia painting Rogier the context of a survey of the scholarship of the day. returned to the same awkward devised for the man footing supporting Christ's legs 4"J.LASSAIGNE: Flemish Painting. The Centuryof van Eyck, Geneva [1957], p.93. in the Descent- left heel legs crossed, lifted and right toe pointing out. See also s.N. 25W. R. VALENTINER: Catalogueof a Collectionof Paintingsand someArt Objects,Flemish and BLUM: Invention in the Art of 'Symbolic Rogier van der Weyden', Konsthistorisktidskrift Dutch Paintings,II, Philadelphia [1913] p.13, nos.334 and 335; SKIDELSKY,loc. cit. at where the [1977], p.1 19, multivalence of the Virgin's pose in the Philadelphia paint- note 23 above; RICHARDSON, loc. cit. at note 7 above, p.64. is ing noted. 26PANOFSKY,op. cit. at note 6 above, pp.249 and 285. 23S. SKIDELSKY: 'Two panels by Van der Weyden on Exhibit in Philadelphia', The 681 ROGIER VAN DER WEYDEN'S PHILADELPHIA 'CRUCIFIXION'

'the treatmentof form and movement are far from abstract reliable indications of the work's pre-1941 appearance.30 and ascetic,' and in which the 'flesh tones and soft colors of Largerareas of retouchingwere carriedout in a modulated the garments create a harmony no less rich for being sub- rigatinotechnique of fine vertical strokesthat makes restora- dued'.27This contrastbetween settingand figureswas in truth tions distinguishablefrom intact originalpaint on close view- a productof restoration.The originaltreatment of the scene, ing, reducing the potential for confusion over the extent of if comparativelyaustere, can now be seen as more generally reconstructionin passagesmost damaged by the 1941 clean- and uniformly consistentwith the conventionsof fifteenth- ing. century Netherlandishillusionism in creating a convincing The spuriousgold backgroundposed an exceptionalprob- perceptualcontinuity between the holy settingof the painting lem. At no time was its removalconsidered justifiable, as no and surroundingsfamiliar to the viewer.28 furtheroriginal surface beyond the small scatteredfragments As the examinationrevealed the extent and natureof dam- of backgroundpaint alreadyfound stood to be revealed.The ages caused by the 1941 cleaning,it also broughtto light dif- gold also marked a significant,if unfortunate,phase of the ficulties arising specificallyfrom the restorationcarried out painting'shistory and, even with the care taken in examining by Rosen. The painting was marredby the uneven applica- the area in 1990, the possibility remained that the gilding tion and pronounced discoloration of the 1941 varnish could yet conceal information of future interest. The gold (alreadyprompting its removal from the figures of St John was neverthelessan essentialfalsification, imposing an unac- and the Virginin 1962)but the greaterproblem was the 1941 ceptable barrierto comprehensionof some of the most basic retouching.Though it dealt tentativelywith even the most aims and achievementsof the work. By its transformationof prominentdisruptions of surfaceand form, it was coarse,and the original dark, recessive background into a bright, flat, particularlyconspicuous where seen alongside the superb reflective surface, the 1941 gilding worked aggressively craftsmanshipof areas of preservedoriginal surface.29 againstthe image, to a degree scarcelyimaginable until a pre- The Crucifixionwas cleaned and restoredin 1992-93. The liminarymasking-out of the gold showed how obviouslythe removalof even merely suspectedrepaints in the 1941 clean- painting'scolour and modelling had been developed by the ing meant that the varnishand retouchingapplied then were artistin referenceto an overalllower value scale. Evaluation virtuallythe only restorationmaterials present.Just fifty years of a range of possibilitiesover a period of months showedthat old, the restorationswere readilydistinguishable and, though a carefullyadjusted facsimile of the former darksky, applied their removal from paint surfacesrendered fragile in places over the gold, most suitablysuggested its pre-1941 appear- by the previouscleaning requiredtime and care, it presented ance and effect. As carried out, the restorationof the sky no extraordinarycomplications. The 1992 cleaning made duplicatesthe colour,texture and thicknessseen in the frag- fully apparentthe widely differingstates of variouspassages, ments of the originalbackground and in comparablefields of presenting a hierarchyof problems to be addressedin the azurite found in other paintingsof the period, but can easily subsequenteffort to reinstate a sense of the painting'spre- be removed to reveal the 1941 gold if any future generation 1941 unity of effect. The approach of the 1993 restoration desires(Fig.23). was shapedby recognitionof two principaldeficiencies of the Due to circumstancesof pre- and post-War scholarship preceding one: its lack of factualjustification for elements and the drasticchange in the stateof the paintingin the inter- introducedby the restorerand its failureto lessen the likeli- vening 1941 restoration,a number of authenticfeatures that hood of the sort of misreadingof incidentaldamage as inten- lent unique distinctionto the PhiladelphiaCrucffixion have not tional artistic effect that is seen in the post-Warliterature. received scholarlyconsideration. Though the painting was Therefore, a primary aim of the 1993 treatment was the accepted and generally praised by eminent scholarsbefore restorationof damaged elements of the complex of known 1939, none provideda detailedanalysis of its stylisticcharac- originalformal qualities,but only to the extent that the pre- teristicsor specific supportfor their appraisalsof its quality. cise characterof those qualitiescould be establishedfrom the The comparativelysuperficial descriptions of the time are combined technicaland documentaryevidence. Reconstruc- due at least in part to the fact that the painting,which has left tive retouching,carried out where volume, relief,colour and Philadelphiaonly once since 1906,31was not readily access- compositional equilibrium were most disrupted, was taken ible to the European scholars whose opinions were most only as far as could be substantiated by close reference to all influential. Their reliance on black-and-white photographs

his 7"Ibid.,p.285. Richardson and then Panofsky both considered the painting's full and Richard Offner as mentors in matters of restoration. King explained that insti- to sophisticated colouring as one of its principal strengths. The view of the panels as tution was in principle sympathetic to Offner's categorical opposition retouching was the was for grisaille or 'demi-grisaille', traceable to Friedlander's inaccurate early description, (' ... that once alteration of the original accepted way open easy could not persists in the literature; however, it is difficult to believe that anyone viewing the compromise, invention and license'). In practice, however, King accept him to be 'uncom- painting directly with an unprejudiced eye would be inclined to such a description. leaving damage fully exposed and deferred to Rosen, believing work of The draperies, possibly in deference to Carthusian taste (see JOLLY,loc. cit. at note 20 promising in his insistence on the preservation of [the damaged art's] origi- Western Acts the above, p. 119) are pale yet pure tints, the Virgin's shading to a fully saturated deep nal character'. See Problemsof the 19th and 20th Centuries,Studies in Art, of and to a burnt but unlike the sort TwentiethInternational the Princeton blue StJohn's deep crimson, ofdemi-grisaille seen, Congressof History ofArt, [1963], IV, pp.152-62, for example, on the exterior of the Ghent altar-piece, the setting is dominated by 184-85). large areas of varied and strong colour, and even the wall of grey stone was enlivened 30The 1993 restoration was substantiated point-for-point by comparisons between the with glazes and strewn with highly coloured detail. the painting's surface, the early photographs and effects seen in other works by at the another late 28JOLLY,loc. cit. at note 20 above, p. 120. If, asJolly suggests, the Philadelphia Crucfix- artist and shop. For example, the large Crucifixion Escorial, paint- in the ion was painted for a Carthusian monastery, the wall's stonework, weathering, ing of large scale and similarly broad design, exhibits points of handling lichens and mosses would have evoked the enclosed exterior spaces familiar to that foreground directly comparable to that seen in the early photographs of the order. Philadelphia painting and indicated by preserved paint in otherwise damaged areas. of "Rosen's minimal yet wilfully apparent retouching was accepted by some at the time Though less directly applicable to the Philadelphia restoration, first-hand study as a logical compromise between the extremes of leaving damage completely visible the meticulously detailed and beautifully preserved Prado Descentfromthe Cross(being of and restoring to the point of falsification. In a 1961 panel discussion on 'The Aes- restored by Maite Davile during the same period) afforded a fuller appreciation thetic and Historical Aspects of the Presentation of Damaged Pictures', Edward S. the range of Rogier's style over the decades. Art Museum in 1939 for King, Director of the Walters Art Gallery, Baltimore, invoked the names of Rosen 3'The painting was lent to the Worcester (Massachusetts) (who had founded the Gallery's technical laboratory in 1934) and fellow-participant the Worcester-Philadelphia Exhibition of Flemish Paintings. 682 ROGIER VAN DER WEYDEN'S PHILADELPHIA 'CRUCIFIXION'

r ir: I~ rlj ;ii

ia ~fXl a~41

L~Yalll al~~SI

?, ?i~~i?-;;": I~ ?:~-:s~sRe~eauls~ LC?IL~l~d? ":

aii)~ -,n C ~I r"@;~

a I -:

~n~

i r

23. The Crucifixionwith the Virgin and St John after the 1993 restoration. permitted comment on little more than general design and torted by the 1941 restoration. From then until 1990 it was made critical evaluation of other (particularly Friedlander's) never suspected that what were considered to be two of the opinions impossible.32When in 1939 the scholarship of the painting's more notable attributes - the gold background and preceding decades was finally challenged, it was precisely an exceeding starkness - were products of restoration that because E.P. Richardson could not help noting the striking weakened or directly conflicted with authentic qualities. The disparity between the painting's superb quality, colour and understanding that is now possible of the visual and material completeness viewed first hand and the prevailing scholarly character of a number of original features removed or dimin- conception of the panels as near-grisaille exteriors of altar- ished in 1941 provides a factual basis for further analysis of piece shutters, suggesting shop production, or an incomplete their broader stylistic implications and their part in the for- triptych. After the War, Richardson was joined in his enthu- mal and textual coherence of the work. The rediscovery of siasm for the painting by Panofsky and others, as the study of fundamental aspects of a major work's strengths and refine- Netherlandish painting was entering a phase of new comment, lost to scholars for five decades, presents an opportuni- plexity in the analysis of style, iconography and technique, ty for a truer measure of the boundaries of Rogier's ambitions but by that time the object of analysis itself was materially dis- and accomplishment. PhiladelphiaMuseum ofArt

"2Anotherimportant quality of the Philadelphia painting that could not be appreci- able to say anythingmore exact [aboutits date] with assurance'(Der Meister von Fld- ated in the study of photographs alone is its imposing scale, witness Friedrich Win- malleundRogier van der Weyden, Strassburg [1913], p.51). The figuresare well over half kler's strange statement that '[the Philadelphia painting] has too small figures to be life-size. 683