Caught Between Dispensations

journal of visual cult ure Caught bet ween Dispensat ions: Het erogeneit y in Early Net herlandish Paint ing Amy Powell Abst ract This article considers the generic and temporal heterogeneity of a painting titled Ho ly Fa mily, made by the Antwerp painter Joos van Cleve around 1515. Hans Belting has used Joos’s hybrid painting, in which a still life appears before the Virgin and Child, to illustrate where ‘the era before art’ ends and the era of art proper begins, placing the genre of religious icon and the genre of still life on opposing sides of the line separating the medieval from the modern world. It is to Belting’s compelling account of the painting that this article responds by asking what is lost in the drawing of such periodizing lines. Keywords disguised symbolism ● early Netherlandish painting ● icon ● periodization ● still life ● uncanny ● period eye Periodization, the parceling of the history of images into a series of discreet epochs, can make certain differences visible, but it exacts a certain price for doing so. Images that sit on the cusp between historical periods have a way of revealing precisely what that price is. Toward the end of Lik eness a nd Presence: The Ima ge Befo re the Era o f Art (1994), Hans Belting discusses an early Netherlandish panel painting, which he describes as half medieval icon and half modern still life. The painting was made by the Antwerp painter Joos van Cleve sometime during the second decade of the 15th century and is titled Ho ly Fa mily (Figure 1). 1 It pictures the Virgin and Child with Joseph standing behind them and a still life of fruit, nuts, a beaker of wine, a platter, and a knife on a ledge before them. In Belting’s reading, half the painting is governed by an old symbolism and half by a new mimesis: journal of visual cult ure [http://vcu.sagepub.com] Copyright © 2008 SAGE (Los Angeles, London, New Delhi and Singapore) Vol 7(1): 83–101 [1470-4129(200804)7:1]10.1177/1470412907087203 Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 84 journal of visual culture 7(1) The still life is the mirror image of natural objects in the sense of the Augustinian a equa lita s, whereas the figure of the Virgin, as an ima go , is a likeness in a different, symbolic sense: it is a preexisting image, not the copy of something from nature. Joos van Cleve distinguished the ‘image’ of the Madonna from the painted still life by setting it at a distance and by quoting the figure from an old painting that symbolically represents the true prototype: in this case it is Jan van Eyck’s Lucca Madonna in Frankfurt [Figure 2]. . . . The former icon appears as a quotation within the modern invention of the artist. (Belting, 1994: 475) Figure 1 Joos van Cleve, Ho ly Fa mily, c. 1515, oil on wood, 16 3 ⁄4 –12 1 ⁄2 in. (42.5 cm–31.8 cm). The Metropolitan Museum of Art, New York, The Friedsam Collection, Bequest of Michael Friedsam, 1931 (32.100.57). Image © The Metropolitan Museum of Art. Reproduced with permission. Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions Figure 2 Jan van Eyck, Ma d o nna a nd Child (Lucca Ma d o nna ) , c. 1434, oil on wood, 26 in.–19 3 ⁄4 in. (66 cm–50 cm). Städelsches Kunstinstitut, Frankfurt am Main, Germany. Photo credit: Foto Marburg/Art Resource, NY. Reproduced with permission. Belting’s interpretation of Joos van Cleve’s painting places it on the cusp between the medieval and the modern worlds, straddling the ‘era before art’ and the era of art proper. This is a compelling description of the painting given that at about 1515 a glance backward would have encountered many icons and only a smattering of pictures that could be called still lifes, a glance forward a gradual replacement of icons with still lifes and other new genres. Janus-faced, the Ho ly Fa mily seems to look in both these directions at once. Because the temporal and generic hybridity of this painting is something it shares with others of its period, Belting’s discussion of the Ho ly Fa mily Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 85 86 journal of visual culture 7(1) dovetails with the most significant critical debate about early Netherlandish painting to have taken place in the 20th century – a debate that concerns the appearance side-by-side of Virgins and things, divine persons in domestic spaces surrounded by prosaic objects. Erwin Panofsky’s theory of disguised symbolism, which he formulated most fully in his Ea rly Netherla nd ish Pa inting: Its Origins a nd Cha ra cter of 1953, and which has become the touchstone of this debate, was meant to explain the mechanism by which Jan van Eyck and his coevals reconciled these irreconcilables: 2 There could be no direct transition from St. Bonaventure’s definition of a picture as that which ‘instructs, arouses pious emotions and awakens memories’ to Zola’s definition of a picture as ‘un coin de la nature vu à travers un tempérament.’ A way had to be found to reconcile the new naturalism with a thousand years of Christian tradition; and this attempt resulted in what may be termed concealed or disguised symbolism as opposed to open or obvious symbolism. (Panofsky, 1953, Vol. 1: 141) For all that it has been justifiably criticized, Panofsky’s theory of disguised symbolism has the considerable merit of reflecting the heterogeneity of early Netherlandish paintings. 3 Where Panofsky’s theory posits that these paintings have a ‘preconceived symbolical program’ that controls their iconography down to the ‘smallest detail’, it is too tidy, but where on the contrary it registers their inconsistencies, it is, it seems to me, extremely useful (p. 137). Panofsky begins his discussion of disguised symbolism by analyzing several paintings of the annunciation, in each of which he identifies Romanesque architecture as a symbol of the old era of Jewish law and Gothic architecture as a symbol of the new era of grace brought about by the incarnation of Christ (pp. 131–40). In these annunciations, architectural symbolism represents ‘the doctrine that the Virgin’s impregnation with the Holy Spirit marks the transition from the Old Dispensation to the New’ (p. 132). The choice of the annunciation as first example in a discussion of disguised symbolism can hardly be arbitrary. It is because of the incarnation that the Christian God is visible in the world. In using the annunciation to illustrate his argument, Panofsky implies that the phenomenon of disguised meaning is not restricted to pictures. It is a mode of symbolism that belongs first to the world (p. 142). But the wo rld of disguised symbolism is a world caught between two dispensations. In none of the annunciations Panofsky considers is the old dispensation, which is represented by Romanesque architectural elements, fully surmounted. In an annunciation attributed to Petrus Christus, ‘the shrine of Judaeo-Christian religion, divided though it is into the Old and New Dispensation (or, as the schoolmen put it, the spheres of ‘imperfect’ and ‘perfect’ revelation), appears as one indestructible edifice’ (p. 134) (Figure 3). 4 This is a world marked by both the old blindness and the new revelation; it is a place in which God both does and does not appear. Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions Figure 3 Attributed to Petrus Christus, Annuncia tio n , c. 1450, oil on wood, 31 in.–25 7 ⁄8 in. (78.7 cm–65.7 cm). The Metropolitan Museum of Art, New York, The Friedsam Collection, Bequest of Michael Friedsam, 1931 (32.100.35). Image © The Metropolitan Museum of Art. Reproduced with permission. Eight years before the publication of Ea rly Netherla nd ish Pa inting, Meyer Schapiro wrote an essay outlining a theory of disguised symbolism much like Panofsky’s. The essay focuses on a mousetrap found in the right wing of an annunciation altarpiece in the Cloisters Collection of the Metropolitan Museum of Art (Figure 4). Schapiro argues that this humble domestic object, which Joseph is busy making as the angel Gabriel appears to Mary in the next room, is a metaphor for the cross on which Christ was crucified. He quotes Augustine and others to demonstrate by what logical feats theologians made the one wooden object stand in for the other: Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 87 88 journal of visual culture 7(1) Figure 4 Robert Campin and assistant, Annuncia tio n Triptych (Méro d e Triptych) , c. 1425, oil on wood, 25 3 ⁄8 in.–46 3 ⁄8 in. (64.5 cm–117.8 cm). The Metropolitan Museum of Art, New York, The Cloisters Collection, 1956 (56.70). Image © The Metropolitan Museum of Art. Reproduced with permission. The devil exulted when Christ died, but by this very death of Christ the devil was vanquished, as if he had swallowed the bait in the mousetrap . . . The cross of the Lord was the devil’s mousetrap; the bait by which he was caught was the Lord’s death. (Augustine quoted in Schapiro, 1979[1945]: 1) The mousetrap qua cross is already a strange thing, but Schapiro makes it stranger still by arguing that, in addition to its theological meaning, it has latent sexual content: the mousetrap is, ‘at the same time, a rich condensation of symbols of the diabolical and the erotic and their repression; the trap is both a female object and the means of destroying sexual temptation’ (p. 9). As does Panofsky, Schapiro sees in early Netherlandish paintings signs of a difficult transition, in which old and new modes of thinking compete for the domain of picture making, neither fully prevailing over the other: The devoted rendering of the objects of the home and the vocation foretells the disengagement of still life as a fully secular sphere of the intimate and the manipulable. Religious thought tries to appropriate all this for itself; it seeks to stamp the freshly discovered world with its own categories, to spiritualize it and incorporate it within a scheme of otherworldly values. (p. 9) Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions Schapiro here identifies a watershed in the history of western painting, but again – as in Panofsky’s discussion of the annunciations – the appropriateness of Schapiro’s central example connects his theory of disguised symbolism to a world beyond painting. The mousetrap is not just any disguised symbol; it is a disguised symbol that symbolizes nothing less than disguise, the ultimate disguise: a God in human form. Christ is as much a disguised symbol as is the mousetrap and the viewer’s failure to recognize the mousetrap as a symbol of redemption is tantamount to the devil’s failure to recognize the crucifixion as trap. The hybridity of early Netherlandish paintings reflects not only the overlapping of art historical periods explicitly identified by Schapiro and Panofsky but also the coexistence of dispensations that is implied in their choice of primary examples. For a world steeped in an incarnation theory that permits the simultaneous availability and unavailability of God to human sight, disguised symbolism was more than clever picture-making. As long as God’s presence in the world was simultaneously as tangible as an object and as intangible as spirit, the disguised symbol was able to generate the uncanniness necessary to make it compelling. Joos van Cleve’s Holy Family, c. 1515 Such a world betwixt and between is the one in which Joos van Cleve’s Ho ly Fa mily appeared and it is, therefore, fitting that Belting should describe it as a bifurcated painting. The evidence Belting (1994) provides for his claim that the Ho ly Fa mily is divided along the lines of genre and period lies beyond the frame of the painting in, on the one hand, Jan van Eyck’s Lucca Ma d o nna and, on the other hand, the world of real objects. For Belting, the difference between these two referents determines the distinction between the icon and the still life. But visual evidence from within Joos van Cleve’s painting can also be marshaled in support of Belting’s claim, and should be in order to make clear that what we are dealing with here is the ‘sensible substance’ of the painting as much as its modes of signification. 5 The still life casts no shadow on the figures behind it, and the figures no shadow on the still life. This tear in the connective tissue of penumbral forms – which binds mother to child and fruit to platter – separates the space of the icon from that of the still life. Although it contains a milky version of the red cherries and the yellowish pomegranate and quince (?), the rim of the shiny, pewter platter likewise fails to reflect the Virgin’s red garment. Color also marks a division between the two spheres. The green of the ledge reappears in the grapes and pear, making it the dominant color in the foreground, while it remains absent from the rest of the painting, except for the moss green of Joseph’s shirt. Finally, we view the ledge and the still life from a wellelevated vantage point that is incommensurate with the angle from which we view the shelf mounted on the wall behind the figures; that shelf reveals its underside to our gaze. This shift in perspective is typical of paintings of the period. Here its effect is to weaken the connection between the two spaces. Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 89 90 journal of visual culture 7(1) If one overlooks the divide that Joos van Cleve inserted between the icon and the still life, one is liable to miss some of the painting’s rhetorical subtlety. The still life has sometimes been read as a translation and elaboration of the first phrase that appears in the scroll in Joseph’s hands, ‘and blessed is the fruit of your womb’ (Luke 1: 42) (Bedaux, 1987: 162). It has also been explained as a complex set of Christological symbols: In this Ho ly Fa mily there is an apple of original sin on the dish . . . the mystic grape; the half of a walnut, exposing the sweet kernel, is the divine nature of Christ, and its shell signifies the wood of the Cross, lignum crucis. From these we pass on to the beaker containing wine, which is a symbol of Christ’s blood poured out on the Cross. On the dish is a pomegranate, typifying the Resurrection, and cherries, specifically regarded as fruits signifying Paradise. (Bergström, 1955: 304) For all its richness, this reading does not address the seemingly accidental arrangement of these objects, the way in which the cherries are scattered with their stems pointing in different directions. A passage from the autobiography of the 14th-century Dominican mystic Heinrich Suso concerning the eating of fruit offers a foil for the painted still life. Suso transforms every cut and each act of consumption into a symbol of someone or something holy: Large pieces of fruit he divided into four parts. Three parts he ate in the name of the Holy Trinity, the fourth part in the love with which the heavenly Mother gave her tender child Jesus an apple to eat. This part he would eat unpeeled because children usually eat it unpeeled. (Suso, 1989: 77) 6 At Suso’s table, no wedge of fruit or scrap of peel is left unaccounted for and objects are arranged – the reader is led to imagine – meticulously. The iconographer is always in danger of reading Joos van Cleve’s still life as if it were Suso’s, that is to say, as if its symbolism were ‘open or obvious’ (Panofsky, 1953, Vol. I: 141), disregarding the painter’s efforts to arrange his fruit seemingly without rhyme or reason. This arbitrariness is what Belting has called the attempt of the still life to ‘deceive the eye by duplicating real objects’; it is what Roland Barthes called the reality effect (Barthes, 1986[1967–80]: 148); 7 and it is the aspect of early Netherlandish painting that provoked Panofsky and Schapiro to concoct a theory unlikely enough to allow a mousetrap to be a device for trapping mice, the cross on which Christ was crucified, and a vagina. Belting’s attention to the novel realism of the still life is a necessary antidote to readings that would turn Joos van Cleve’s table into Suso’s. 8 But his recognition that the icon and the still life demand different modes of reading should be a precursor to the question of what to make of the fact that, despite their differences, they do indeed appear together in this painting, and they do appear very much to gether. The icon, which is a quotation from Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions Jan van Eyck’s Lucca Ma d o nna of around 1434, lacks quotation marks. If Joos van Cleve had wanted a pictorial equivalent to the quotation mark, something alerting us to the fact that the icon has a status different from that of the still life, he might have used angels or a halo to indicate that the latter is a vision. By leaving these out, he brings the icon and the still life into the greatest possible intimacy without, that is, allowing them to completely merge with one another. Two objects reinforce this intimacy. The piece of fruit cradled in the hand of the Christ child, embedded in the hollow between his belly and the body of the Virgin, seems to have strayed from the world of the still life. With one hand on his mother’s wrist and one cupped around the stolen piece of fruit, the Christ child cross-fertilizes the two worlds. The beaker of wine in the lower left corner also pulls the two parts of the painting together. Its transparent glass reveals the cloth of the Virgin’s robe behind it. To emphasize this overlap, the painter made the folds of the cloth run parallel to the right edge of the beaker and parallel again to the three highlights (a Trinitarian symbol?) that fall on the glass. These lines are the culmination of the long diagonal that extends from the Virgin’s chin along her breast, through the sightline of the Christ child and down the length of the Virgin’s red robe. The wine of the Passion held in the beaker seems to have flowed along this trajectory, a distillation of the living couple of mother and child. The lone piece of fruit cradled in the Christ child’s hand and the red of the Virgin’s robe distilled into the liquid in the beaker are the elements that cross over from the icon to the still life and back again, transgressing the division between the two and complicating Belting’s periodizing efforts in the process. To see this painting, one needs to see both the difference between the icon and the still life identified by Belting and the fact that the painting brings them together, and then one needs to ask how in their proximity they operate on one another. The St ill Life When the still life and icon are seen together in their difference, the icon inflects the fruit, not only by lending it a set of symbolical meanings as some have argued, but also by turning it into a metaphor for the picture’s capacity to produce meaning abundantly. 9 In the late middle ages, scripture was described as fecund and compared to fruit and fruit trees, and the process of its exegesis was metaphorically designated as a harvesting or pressing of fruit. In his monumental work, The Fo ur Senses o f Scripture, Henri de Lubac collected examples of the trope of scriptural fecundity, for example, as it appears in a letter of 1452 written by Nicholas of Cusa: ‘The inexplicable fecundity of Divine Scripture is diversely explained by diverse writers, so that its infinity might shine forth variously in a great number of ways’ (Nicholas of Cusa, quoted in Lubac, 1998, Vol. 1, 328, no. 44). 10 A 12th-century abbot, perhaps Gilbert of Hoyland, employs the metaphor: Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 91 92 journal of visual culture 7(1) For you know that the words of Sacred Scripture, when they have been gleaned for a first and second vintage, can be squeezed so as to yield a rich and intoxicating liquid, as if it were usual for them, laden with sacred meanings as they are, to flow again even as if they had never been touched in the first place. 11 In his Epistles, the 12th-century, Praemonstratensian abbot Philip of Harvegnt explains: Though a manifold mystery is contained in the things that have been said, that mystery is not held to be contrary to the literal sense . . . one arrives the more gracefully at the innermost fruit by as much as the surface of the letter is found to be not only leafy with words but also fruitful with truth. 12 The trope of the fecundity of scripture as these late medieval writers elaborate it pertains specifically to scripture in its relation to readers and to what is made available through their interpretive efforts. The personal address of scripture to the reader was the subject of what medieval exegetical terminology designated as the tropological mode of reading. Lubac explains this mode: In the most general acceptation, a trope was a figure, a mode, or a turn of phrase ([Gk.] tro po s, [Lat.] co nversio ), by which one turns some expression to designate some object other than the one naturally meant. Tro po lo gia , accordingly, was a ‘speech turned around’ or ‘turning’ something else ‘around’; it was a ‘turned’ or ‘turning manner of speech’. 13 (Lubac, 1998, Vol. II: 129, see also 127) Medieval exegetes then transposed the turning within the text onto the relationship between the reader and its text: ‘[Tropology] would be explained as being “a speech turned around toward us, i.e., toward our ways of behaving,” or “a speech turning around, pertaining to the mind’s ways of behaving”’ (Lubac, 1998, Vol. II: 129). 14 In its ‘turning manner of speech’, tropology finds in scripture the personal address to the reader, discovering where scripture speaks as if it knew that this or that particular reader were there. Quoting Saint Gregory, Lubac writes: It is by the tropological sense thus understood that Scripture is fully fo r us the Word of God, this Word which is addressed to each person, hic et nunc [‘here and now’] as well as to the whole Church, and telling each ‘that which is of interest to his life’. (Lubac, 1998, Vol. II: 140) In being for us, regardless of who, when, or where we are, the text is volatile. By providing lessons for ‘our way of behaving’, scripture changes along with the readers before whose eyes it is brought. The capacity of scripture to enter the body and undermine the distinction between text and reader then finds Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions its metaphor in a fruit that is comestible. For Alan of Lille, ‘the flower of allegory gives off its scent, the fruit of tropology suffices’. 15 Tropology’s turning toward the reader becomes a turning into the reader through ingestion. Understood in an analogous manner, the fruit in Joos van Cleve’s Ho ly Fa mily not only symbolizes specific elements of the narrative of Christ’s life, but also functions as a metaphor for the capacity of the image to produce meaning – as scripture does – abundantly and as if in response to individual viewers. The location of the still life on this side of the ledge reinforces the sense that it is for us. The knife and platter are pushed so far in our direction that they threaten to fall into our space; the glass and nuts are also here on our side of the table. The painting is not a mirror that once reflected the world in which it was made – as most definitions of realism would have it – but rather something uncanny in its ever-renewable capacity to reflect the world in which it no w exists, uncanny insofar as the uncanny ‘is often and easily produced when the distinction between imagination and reality is effaced’ (Freud, 1953[1919]: 244). Where Suso’s apples can be fully explicated through a symbolic decoding, Joos van Cleve’s fruit is more complex, containing in its delicate flesh a life vibrating in response to the viewer’s imagination. The complexity of Joos van Cleve’s fruit gets overlooked if one follows Belting in rendering the icon and the still life each to its own discreet world. To read the still life tropologically, that is, to see it as an uncanny address to the viewer on the model of scripture, requires seeing across periods. Belting’s periodization may well be indispensable to understanding the difference between these two modes of painting, but when followed too rigorously it prevents one from seeing what their proximity precipitates (see Damisch, 1997: 82–3). The disciplinary quarantining of ‘old’ and ‘new’ kills the fleeting sense of strangeness that comes from their meeting. Against the practice of periodization, Georges Didi-Huberman has insisted on a mode of art history that is sensitive to hybridity: ‘The history of images is a history of objects that are temporally impure, complex, overdetermined. It is therefore a history of polychronistic, heterochronistic, or anachronistic objects’ (DidiHuberman, 2003: 42; see also Nagel and Wood, 2005). Historical images are already – before the intervention of the historian whose relationship to the image is by definition anachronistic – temporally complex and, in the case of the Ho ly Fa mily and other early Netherlandish paintings, generically and ontologically complex as well. The Icon If reassembling the pieces of Joos van Cleve’s painting invests the still life with an uncanny volatility and sensitivity to the viewer’s imagination, it also has its effect on the icon. Belting explains Joos van Cleve’s quotation from Jan van Eyck’s Lucca Ma d o nna as a tapping of the authority of an older model. This he views as a medieval strategy: ‘For the medieval mind, the old Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 93 94 journal of visual culture 7(1) was the mark of tradition, which in turn guaranteed authenticity, the rampart against the flood of time. This idea was also applied to images’ (Belting, 1994: 432). But the exactness of early Netherlandish copies at 1500 distinguishes them from their medieval predecessors, which recapitulated familiar patterns and motifs but tended to do so more loosely. 16 In addition to their exactness, early Netherlandish copies are remarkable for their tendency to reproduce works from only a generation or so earlier and from their own tradition. Only about 80 years and the distance between Antwerp and Bruges, for example, separate Joos van Cleve’s Ho ly Fa mily from van Eyck’s Lucca Ma d o nna . The self-reflexivity of early Netherlandish copying led Wolfgang Schöne to speak of a ‘narcissism of Netherlandish art’ (Schöne, 1938: 72). Panofsky calls archaism at 1500 a form of nostalgia: ‘The works of the great masters came to be looked upon “at a distance”: no longer as possessions, however cherished, but as monuments of a past worthy of study and admired with a feeling akin to nostalgia’ (Panofsky, 1953: 351). Paul Philippot stresses the newness of the aesthetic attitude toward the image evinced by early Netherlandish copying: ‘The becoming conscious of the image as image is also at the origin of a phenomenon characteristic of the end of the fifteenth century in the southern Low Countries: exact copies’ (Philippot, 2005: 174). Schöne, Panofsky, and Philippot all consider archaism at 1500 a modern phenomenon, whereas Belting seems to view it primarily, if not exclusively, as a perpetuation of medieval practices. 17 This difference notwithstanding, each of the four grapples with the question of whether archaism is a product of inclusion or exclusion (proximity or distance) vis-à-vis tradition, and each represents the relation of painting at 1500 to its history as a form of objectification. Already in 1956, when Johannes Taubert took up the issue of early Netherlandish copying, he had brought this line of thinking to its logical conclusion, arguing that the realism of early Netherlandish painting conditioned artists to see the world as a collection of things, and that it was this mode of seeing that they directed toward paintings, such that paintings appeared as objects like any other objects in the world. (Taubert, 2003: 140). This making of a painting into an object is precisely what happens in Joos van Cleve’s Ho ly Fa mily, where the objectification of the icon that is initiated through quotation is reinforced by the presence of the still life in front of it. When placed behind the still life, the quotation of the old icon becomes less a claim to authority – as Belting suggests – than a conflation of the animate and the inanimate. ‘Doubts whether an apparently animate being is really alive; or conversely, whether a lifeless object might not be in fact animate’ give rise, according to Freud, to uncanny effects (Freud, 1953[1919]: 226, quoting Ernst Jentsch). The human figures in this painting, although convincingly three-dimensional, are indistinguishable in their stillness from the objects on the ledge. The proximity of the living mother and child to the collection of inanimate objects makes the human figures, already reified through quotation, strangely lifeless. Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions If we follow Belting in seeing the icon in a register different from that of the still life and also in seeing it sepa ra tely from the still life, we can restore to it a certain degree of liveliness. This requires strictly differentiating between the mimetic representation of objects in the world (art) and the quotation of the icon (the image before the era of art). But for many who have commented on early Netherlandish copying and quotation – including Schöne, Taubert, Panofsky, and Philippot – this distinction is what early Netherlandish copying calls into question. For each of these commentators, that copying entails an objectification of the old image that transforms it into a thing difficult to distinguish from the furniture and fruit with which Joos van Cleve so tellingly brings it together. It is only by seeing the two parts of Joos van Cleve’s painting to gether without seeing them as the same, that we can do justice to this uncanny ambiguity. ‘Eyes Here! Eyes Here!’ 18 When questioned about the Sand-Man, his mother, it is true, denied that such a person existed except as a figure of speech; but his nurse could give more definite information: ‘He’s a wicked man who comes when children won’t go to bed, and throws handfuls of sand in their eyes so that they jump out of their heads all bleeding. Then he puts the eyes in a sack and carries them off to the half-moon to feed his children.’ (Freud, ‘The Uncanny’, 1953[1919]: 227–8) For Belting, the two parts of Joos van Cleve’s painting must be separated because they represent two different ways of seeing. The period divisions that structure the discipline of art history have been justified both by a belief in the relativity of vision and by demands for scholarly expertise. They represent perceived watersheds in what, by a certain turn of speech, has come to be called the period eye (see Baxandall, 1988[1972]: 29–108). Periodization ensures that the historian will gather the information necessary to allow him or her to exchange his or her eye for the eye of the period that is his or her subject. Taken literally and against any claim that it is merely a figure of speech, this fantasy of trading eyes is macabre and brings to mind Freud’s reading of E.T.A. Hoffman’s story The Sa nd ma n , in which eyes are snatched, pecked, stolen, sold, supplemented with prosthetics, and constructed entirely artificially – all, Freud tells us, as displaced forms of castration. The concept of the period eye, indeed, checks the fantasy of the omniscient eye, cutting it down to size with the sobering fact of historical difference. In this regard it is salutary, but a two-fold danger lurks in the use to which the concept has often been put. The first is that it can turn the historian’s relationship to historical images into a form of hallucination; the second is that it can rob images of their incoherencies. A crucial moment in Belting’s reading of Joos van Cleve’s Ho ly Fa mily, the sentences devoted to the old man in the background of the painting, concerns directly this problem of eyes. Belting finds in this figure not a Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 95 96 journal of visual culture 7(1) picture of Joseph, whom we would expect, but rather a portrait of the late medieval viewer: The old man who appears in the background before an open window . . . is no longer . . . Joseph from a ‘Holy Family’ but the client who commissioned the painting . . . The figure of the Virgin in the painting is the one that appears before his inner eye, which is why he has taken off his reading glasses. This finally helps to exclude any literal meaning of the image. (Belting, 1994: 275) By removing his glasses, Belting argues, the old man reveals himself to be an embodiment of a peculiarly medieval mode of viewing that has little to do with seeing in the ordinary sense. But why, one might ask, would the old man need to take off his reading glasses in order to behold a vision of the Virgin with his inner eye? Reading glasses do not affect inner eyes. Reading glasses are removed in order to look at something other than writing with one’s ordinary eyes. Belting’s way of thinking about seeing in early Netherlandish painting has a tendency to domesticate its strangeness by relegating what it represents to the category of phantasmagoria. Belting is not alone in this line of thinking. Craig Harbison believes that the canon in Jan van Eyck’s Ma d o nna with Ca no n va n d er Pa ele ‘remov[es] the eye-glasses which he needs for his earthly vision. What he sees before him is not, of course, merely earthly, nor does he see it, strictly speaking, except in his mind’s eye’ (Harbison, 1985: 100; see also Rothstein, 1999) (Figures 5 and 6). But, once again, the glasses removed are reading glasses. The nearsighted canon removes them, I would suggest, to see with his corporeal eyes the Virgin who would otherwise be nothing but a blur. In which case, the removal of the glasses demonstrates that this Virgin has made her appearance in the flesh. This seemingly minor issue of a pair of reading glasses represents the rather major issue of how medieval viewers looked at sacred images. Belting’s interpretation of the old man in the Ho ly Fa mily and Harbison’s interpretation of the canon in the Ma d o nna with Ca no n Va n d er Pa ele amount to an argument that the late medieval viewer did not look at the sacred picture with corporeal eyes but instead used that picture to generate an image before his or her inner eye. This is compelling insofar as the appearance of the Virgin in an ordinary domestic space indeed calls for an explanation of some kind. But my eye fails to find evidence that the donors depicted in the Ho ly Fa mily and the Ma d o nna with Ca no n Va n d er Pa ele see anything other than the world so vividly represented around them – a world that happens miraculously to contain the Virgin. It is the fact that these donors see what cannot be seen that makes these paintings strange. Precisely this strangeness is filtered out by the theory that early Netherlandish paintings represent the hallucinatory visions of their donors. In the case of the Ho ly Fa mily – if the removal of the reading glasses does not settle the matter – the lasciviousness of Joseph’s gaze, which reflects the popular iconography of the ill-matched lovers, in which a wizened old man gropes a young woman, should give Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions 97 Figure 5 Jan van Eyck, Ma d o nna with Ca no n va n d er Pa ele , 1436, oil on wood, 48 in.–61 3 ⁄4 in. (122 cm–157 cm). Groeningemuseum, Bruges, Belgium. Photo credit: Erich Lessing/Art Resource, NY. Reproduced with permission. Figure 6 Detail of Figure 5. Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 98 journal of visual culture 7(1) pause to anyone tempted to imagine that this man is not in the presence of an entirely corporeal woman. Belting’s figure of the medieval donor, who ‘sees’ the image only with his inner eye, is intended to illustrate the fact that that person did not see the way people in the modern world do, but I would also venture to suggest that Belting makes of this figure a model for the conscientious historian of images. That historian is meant to ‘see’ the painting by gazing beyond it to a hallucinatory vision of what it once was, which means looking at the painting as if it were no longer available to sight, but instead only to the imagination and only by means of the labor of gathering up again the historical context that has fallen away from it. According to this dogma, the person who looks too directly at the image makes the idolatrous mistake of worshipping mere matter. The good historian beholds with his or her inner eye the image of the painting, as it really was when it was visible – as it no longer is – to the outer eye. In such a writing of the history of images, the rules of period divisions are of paramount importance and the appearance side-by-side of historically distant images is strictly prohibited; the image is imprisoned in a historical hothouse, shielded from winds that might perchance bring it into contact with something in whose presence it would appear strange, even where that something is to be found within its very own frame. Acknowledgement s Versions of this article were presented at the 2003 Interdisciplinary Graduate Student Conference sponsored by the Humanities Center and English Department of Harvard University and at the Renaissance Colloquium held at Temple University in 2005. For their insights and suggestions, I would like to thank Raiford Guins, Richard Neer, Alicia Walker, Henri Zerner, and the anonymous readers and editorial board of the jo urna l o f visua l culture . Not es 1 On Joos van Cleve, see Hand (2004). On the Ho ly Fa mily, see Ainsworth (1998: 246–9). On the appearance and development of devotional images with still life elements, see Falkenburg (1994: 6–7, 14). 2. For references to earlier theories of disguised symbolism in Panofsky’s own work and in the work of others, see Pächt (1956: 275). 3. For one of the first and most trenchant critiques of Panofsky’s concept of disguised symbolism, see Pächt (1956: 275–9), Benjamin (1976: 11–24) and Bedaux (1986: 5–28). Disguised symbolism is a subcategory of symbolism and its interpretation a subcategory of iconology. Iconology has come in for a great deal of criticism, which I regard as very useful. A particularly good critique of it is to be found in Damisch (2005). 4. In Matthew, Christ speaks of the ongoing validity of the law: ‘Think not that I have come to abolish the law and the prophets; I have come not to abolish them but to fulfil them’ (Mt 5:17). 5. On iconology’s reduction of painting to a system of signification at the expense of its ‘sensible substance’, see Damisch (2005: 264–6). 6. ‘Daz gross ops teilt er in vier teil: dú drú ass er in dem namen der heiligen drivaltekeit, daz vierde teil in der minne, als dú himlesch múter irem zarten Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. kindlin Jesus ein epfelli gab ze essen. Daz selb teil ass er unbeschniten, wan es dú kindelú also unbeschnitten pflegent ze essene’ (Seuse, 1907: 25). On the reality effect in early Netherlandish painting, see Moxey (1996). Belting by no means entirely discounts the idea that the still life has symbolic meaning determined by the story of Christ’s Passion. He simply stresses the fact that its rhetorical mode is one of realism (Belting, 1994: 474–5). For a discussion of the polysemy of late medieval still life, see Falkenburg (1994: 93). ‘Inexplicabilis divine scripture fecunditas per diversos diverse explicatur, ut in varietate tanta eius infinitas clarescat’ (Nicholas van Cusa, quoted in Vansteenberghe, 1920: 111). ‘Nosti enim quod sacrae verba Scripturae, cum semel et bis vindemiata sunt, ac si gravida sacris iterum soleant influere sensibus et quasi prius tacta non essent, ubertem et inebriantem effundant expressa liquorem’ (Gilbert of Hoyland, quoted in Lubac, 1998, Vol. I, 79). For other uses of fruit, nuts, and tables laden with food as metaphors for scriptural fecundity, see Lubac (1998: Vol. I: 29, 62, 75; Vol. II: 68). ‘Esti in his quae dicta sunt mysterium multiplex continentur, illud tamen litterali sensui contrarium non habetur . . . nec tamen ab ea litteralem intelligentiam removeri; et eo gratius ad fructum intimum pervenitur, quo, non solum verbis frondosa, sed et veritate fructuosa litterae superficies invenitur’ (Philip of Harvegnt, quoted in Lubac, 1998, Vol. II: 68, 289, n. 119). The quotations are from Philo, Augustine, Isidore of Seville, Hugh of St. Victor, and Peter Comester. The four-fold structure of medieval exegesis comprised the historical, allegorical, tropological, and anagogical modes of interpretation. The quotations are from Peter Comestor. ‘Redolet flos allegoriae, satiat fructus tropologiae’ (Alan of Lille, quoted in Lubac, 1998, Vol. II: 129). On the scarcity of exact copies in medieval art, see Alexander (1989: 61). On the use of labor-saving devices in medieval art, see Scheller (1995: 70–7). On the novelty of the exactness of early Netherlandish copies, see Taubert (1960: 67). For an overview of the literature on these copies, see Van den Brink (2001). Belting (1994) also alludes to a new consciousness of history reflected in the practice of copying (p. 475). In summarizing E.T.A. Hoffman’s The Sa nd ma n , Freud writes: ‘His father and the guest are at work at a brazier with glowing flames. The little eavesdropper hears Coppelius call out: “Eyes here! Eyes hear!” and betrays himself by screaming aloud’ (Freud, 1953[1919]: 228). References Ainsworth, Maryan W. and Christiansen, Keith (eds) (1998) Fro m Va n Eyck to Bruegel: Ea rly Netherla nd ish Pa inting in the Metro po lita n Museum o f Art. New York: Metropolitan Museum of Art. Alexander, Jonathan (1989) ‘Facsimiles, Copies and Variations: The Relationship to the Model in Medieval and Renaissance European Illuminated Manuscripts’, Stud ies in the Histo ry o f Art 20: 61–72. Barthes, Roland (1986[1967–80]) ‘The Reality Effect’, in The Rustle o f La ngua ge, trans. Richard Howard, pp. 141–8. Berkeley: University of California Press. Baxandall, Michael (1988[1972]) Pa inting a nd Experience in Fifteenth Century Ita ly: A Primer in the So cia l Histo ry o f Picto ria l Style. Oxford: Oxford University Press. Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 99 100 journal of visual culture 7(1) Bedaux, Jan Baptist (1986) ‘The Reality of Symbols: The Question of Disguised Symbolism in Jan van Eyck’s Arno lfini Po rtra it’, Simio lus 16: 5–28. Bedaux, Jan Baptist (1987) ‘Fruit and Fertility: Fruit Symbolism in Netherlandish Portraiture of the Sixteenth and Seventeenth Centuries’, Simio lus 17: 150–68. Belting, Hans (1994) Lik eness a nd Presence: A Histo ry o f the Ima ge befo re the Era o f Art, trans. Edmund Jephcott. Chicago: University of Chicago Press. Benjamin, Lloyd W. (1976) ‘Disguised Symbolism Exposed and the History of Early Netherlandish Painting’, Stud ies in Ico no gra phy 2: 11–24. Bergstöm, Ingvar (1955) ‘Disguised Symbolism in ‘Ma d o nna ’ Pictures a nd Still Life’, Burlingto n Ma ga zine 97: 303–8. Damisch, Hubert (1997) Mo ves: Pla ying Chess a nd Ca rd s with the Museum . Rotterdam: Museum Boijmans Van Beuningen. Damisch, Hubert (2005) ‘Eight Theses For (or Against?) a Semiology of Painting’, Oxfo rd Art Jo urna l 28: 257–67. Didi-Huberman, Georges (2003) ‘Before the Image, Before Time: The Sovereignty of Anachronism’, in Claire Farago and Robert Zwijnenberg (eds) Co m pelling Visua lity: The Wo rk o f Art in a nd o ut o f Histo ry, pp. 31–44. Minneapolis: University of Minnesota Press. Falkenburg, Reindert L. (1994) The Fruit o f Devo tio n: Mysticism a nd the Ima gery o f Lo ve in Flemish Pa intings o f the Virgin a nd Child , 1450–1550, trans. Sammy Herman. Amsterdam: John Benjamins. Freud, Sigmund (1953[1919]) ‘The Uncanny’, in James Strachey (ed.) The Sta nd a rd Ed itio n o f the Co mplete Psycho lo gica l Wo rk s o f Sigmund Freud vol. 17 , pp. 219–56. London: The Hogarth Press and the Institute of Psycho-Analysis. Hand, John (2004) Jo o s va n Cleve: The Co mplete Pa intings. New Haven, CT: Yale University Press. Harbison, Craig (1985) ‘Visions and Meditations in Early Flemish Painting’, Simio lus 15: 87–118. Lubac, Henri de (1998) Med ieva l Exegesis. Grand Rapids, MI: W.B. Eerdmans. Moxey, Keith (1996) ‘Reading the “Reality Effect”’, in Gerhard Jaritz (ed.) Pictura qua si fictura : Die Ro lle d es Bild es in d er Erfo rschung vo n Allta g und Sa chk ultur d es Mittela lters und d er frühen Neuzeit, pp. 1 5 –2 1 . Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Nagel, Alexander and Wood, Christopher S. (2005) ‘Interventions: Toward a New Model of Renaissance Anachronism’, Art Bulletin 87: 403–15. Pächt, Otto (1956) ‘Panofsky’s “Early Netherlandish Painting”’, Burlingto n Ma ga zine 98: 110–116, 267–279. Panofsky, Erwin (1953) Ea rly Netherla nd ish Pa inting: Its Origins a nd Cha ra cter. Cambridge, MA: Harvard University Press. Philippot, Paul (2005) ‘La fin du XVe siècle et les origines d’une nouvelle conception de l’image dans la peinture des Pays-Bas’, in Ja lo ns po ur une métho d e critique et une histo ire d e l’a rt en Belgique , pp. 153–95. Bruxelles: La Part de l’Oeil. Rothstein, Bret (1999) ‘Vision and Devotion in Jan van Eyck’s Virgin a nd Child with Ca no n Jo ris va n d er Pa ele ’, Wo rd a nd Ima ge 15: 262–76. Schapiro, Meyer (1979[1945]) ‘Muscipula Dia bo li : The Symbolism of the Mérode Altarpiece’, in La te Antique, Ea rly Christia n a nd Med ia eva l Art, pp. 1–19. New York: G. Braziller. Scheller, Robert W. (1995) Exemplum: Mo d el-Bo o k Dra wings a nd the Pra ctice o f Artistic Tra nsmissio n in the Mid d le Ages (ca . 900 – ca . 1470), trans. Michael Hoyle. Amsterdam: Amsterdam University Press. Schöne, Wolfgang (1938) Dieric Bo uts und seine Schule. Berlin: Verlag für Kunstwissenschaft. Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Powell Caught bet ween Dispensat ions Seuse, Heinrich (1907) ‘Seuses Exemplar’, in K. Bihlmeyer (ed.) Deutsche Schriften im Auftra g d er Württembergischen Ko mmissio n für La nd esgeschichte, 1–401. Stuttgart: W. Kohlhammer. Suso, Heinrich (1989) The Exempla r with Two Germa n Sermo ns, trans. Frank Tobin. New York: Paulist Press. Taubert, Johannes (1960) ‘Die beiden Marienaltäre des Rogier van der Weyden: Ein Beitrag zur Kopienkritik’ (The Two Altars of Rogier van der Weyden: A Contribution to the Study of Copies), Pa ntheo n 18: 67–75. Taubert, Johannes ( 2 0 0 3 ) Zur k unstwissenscha ftlichen Auswertung vo n na turwissenscha ftlichen Gemä ld euntersuchungen (Toward an Aesthetic Analysis of the Scientific Examination of Paintings). Munich: Siegl. Van den Brink, Peter (2001) ‘The Art of Copying. Copying and Serial Production in the Low Countries in the Sixteenth and Seventeenth Centuries’, in Brueghel Enterprises, pp. 12–43. Maastricht: Bonnenfanten Museum. Vansteenburghe, Edmond (1920) Auto ur d e la d o cte igno ra nce: une co ntro verse sur la théo lo gie mystique a u XVe siècle . Munster (Westphalia): Aschendorff. Amy Powell is Assistant Professor of Art History at Temple University. She has published on early Netherlandish painting in The Art Bulletin and Art Histo ry, and is currently working on a book about the impact that Protestant iconoclasm had on the genre of landscape as it developed in 17th-century Netherlands. Address: Art History Department, Temple University, 1301 Cecil B. Moore Avenue (004-01), Ritter Annex 857, Philadelphia, PA 19122-6091, USA. [email: amykpowell@ post.harvard.edu] Downloaded from http://vcu.sagepub.com at HARVARD UNIV on March 31, 2008 © 2008 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 101