Richard Woodfield, 'Gombrich, Formalism and the Description of Works of Art' more

Preprint of British Journal of Aesthetics, Vol. 34, No. 2, April, 1994 GOMBRICH, FORMALISM AND THE DESCRIPTION OF WORKS OF ART Richard Woodfield CECIL GOULD once contrasted Kenneth Clark and E. H. Gombrich in the following terms: In looking at works of art Kenneth Clark, a Scotsman brought up in England, started with the visual aspect the relation of form and colour which appealed, or failed to appeal, to his eve and then allowed his mind to consider it. Gombrich, on the other hand, who grew up in the heady intellectual atmosphere of Vienna of the 192os, appears to do the opposite. With him the cerebral element appears to predominate over the visual.1 The reader of Clark's autobiography Another Part of the Wood2 will, however, discover his attachment to the great Viennese historian Alois Riegl: I had read, with immense difficulty, the works of Riegl and had formed the ambition to interpret every scrap of design as the revelation of a state of mind. I dreamed of a great book which would be the successor to Riegl's Spätrömische Kunst-Industrie, and would interpret in human terms the slow, heavy, curve of Egyptian art or the restless, inwardturning line of Scythian gold, and would stand for hours looking at Anglo-Saxon ornament and try to describe how its rhythms differed from the decorations on a Chinese bronze.3 I am sure that there were quite as many connoisseurs in Vienna as there were in England willing to respond to the purely visual appeal of works of art. As we shall see Riegl himself stood in need of a `cerebral' corrective. What is intriguing, however, is Gould's characterization of Clark's perceptual process, which has its roots in the history of formalism. As Gombrich has pointed out, it was Ruskin who gave currency to the idea of the `innocent 4 eye and it was Conrad Fiedler who challenged the idea that the artistic image could escape the mediation of thought.5 Fiedler's formalism reflected his friend, Hans von Marée's, rejection of contemporary interest in literary and anecdotal imagery: `interest in art begins only when interest in literary content vanishes. The content of a work of art that can be grasped conceptually and expressed in verbal terms does not represent the artistic substance which owes its existence to the creative powers of the artist'.6 This did not imply that the work of art would have no subject but that the interest of the subject was purely what was experienced visually. Nevertheless he distinguished between the artist's perceptual experience of the world and the conceptualization of that visual experience in the work of art. The artist's ideal but impossible aim was to arrive at some kind of fit between visual experience and depiction: It is the essential characteristic of the artist's nature to be born with an ability in perceptual comprehension and to use that ability freely. To the artist, perceptual experience is from the beginning an impartial, free activity, which serves no purpose beyond itself and which ends in that purpose. Perceptual experiences alone can lead the artist to artistic configurations. A review of New Light on Old Masters (1986) in Apollo (January 1987), p. 75. Kenneth Clark, Another Part of the Wood (London, 1985). 3 Ibid., p. 108. 4 E. H. Gombrich, Art and Illusion (Oxford, 1986), p. 250. 5 Conrad Fiedler, Über die Beurteilung von Werken der Bildenden Kunst (1876), trans. H. Schaeffer-Simmern & F. Mood, On Judging Works of Visual Art (Berkeley and Los Angeles. 1949). 6 Ibid., pp. 10-11. 2 1 To him the world is but a thing of appearances. He approaches it as a whole and endeavours to recreate it in a visual whole. The essence of the world which he tries to appropriate mentally and to subjugate to himself consists in the visible and tangible configuration of its objects.7 The difficulty the artist continually faced was the intervention of knowledge in perception of the world: `even the simplest sense impression that looks like merely the raw material for the operations of the mind is already a mental fact, and what we call the external world is really the result of a complex psychological process'.8 What Fiedler treated as a difficulty his friend, the neoclassicist sculptor Adolf von Hildebrand, turned to positive account by insisting on the need for clarity and structure in the visual image.9 The depiction of immediate response to sensory stimulus simply resulted in formlessness: `An artistic representation, to be strong and natural, must bring to bear, out of the embarrassing richness which Nature affords, and in spite of it, those elementary effects which correspond to our most general conception of form'.10 He argued that tactile memories played an important role in our recognition of objects. It was the artist's business to create images which would trade on both kinaesthetic and optical sensations. A picture is a static image of a world which we move around in, handle and see with both eyes. The painter's problem was to reconcile this freedom of movement, and point of view, with the need to create a single image. Proxemic and distant vision placed different demands on the spectator: from a great distance relationships became obvious at a glance, closer views demanded a kind of visual compounding activity. According to Hildebrand, the very purpose of a picture was to awaken the idea of space11 but the coherence and unity of a picture was quite distinct from the coherence and unity of nature itself.12 The idea of representation was pivotal to Hildebrand's formalism as while: Ibid., pp. 42-3. Quoted by Gombrich in Art and Illusion (Oxford, 1986), p. 53 from `Moderner Naturalismus und künstlerische Wahrheit', in Schriften über Kunst (Leipzig, 1896), p. 177. In his Durham lecture of 1955, Gombrich remarked `It is an intriguing fact that Hildebrand had made a portrait bust of the psychologist Helmholtz. One would like to know what the two talked about during the sitting' (unpublished ms. p. 3). For comments on the von Marées circle see Kenworth Moffett, Meier-Graefe as art critic (München, 1973), passim. 9 The parallel with the history of Impressionism in France is striking. In 188o, Emile Zola declared that Impressionist art lacked structure and Seurat's painting responded to `formlessness' with a neo-classical severity. It would be interesting to know whether there was a critical network around 'Parisian' Impressionism which extended to Italy. See S. Loevgren, The Genesis of Modernism (London, 1971). 10 Adolf von Hildebrand, Das Problem der Form in der bildenden Kunst (1893), trans. Max Meyer and Robert Morris Ogden, The Problem of Form in Painting and Sculpture (New York, 1907). It is worth noting that Hildebrand defined the problem which Gombrich eventually set out to solve in Art and Illusion: `Considering how different a thing a picture is from the natural object which it represents, the force of the illusion which it produces remains a riddle unless we understand that the picture depends for its spatial effect on no other subjective conditions than does Nature. Neither produces spatial ideas directly, but only indirectly, and that through the medium of the same class of mental processes. Since Nature and the picture both stimulate our sense organs in the same manner, we arrive, in fact, at the same resultant idea. The parallel between Nature and Art is not to be sought in the equality of their actual appearances, but rather that both have the same capacity for producing visual effects. The value of the picture does not depend on the success of a deception, as does the popular value of a panorama, but on the intensity of the unitary spatial suggestiveness concentrated in it'. Ibid., pp. 55-6. 11 Hildebrand, p. 49. 12 Hildebrand, p. 52. This is a notion that Gombrich explored to great effect in his lecture `Raphael's "Madonna della Sedia" reprinted in Norm and Form (London, 1966); he argued the painting was a solution to the problem of reconciling two 'mutually limiting demands that of lifelikeness and that of arrangement' (p. 74). 8 7 `the elements of sensation which compose these spatial values are lines, light and shade, and colors ... it is to be noted that these constitute a spatial value and act on our form conceptions only when regarded by us as representing an object'.13 The artist's forms were recognized as objects in space and the space they inhabited was prior to their depiction: `Ideas suggested by form, as expressive not of space, but of organisation, function or movement, take their place as factors in art only after the spatial ideas are established'.14 While Hildebrand linked his theoretical ideas to a neo-classical sculptural practice his book played an important role in creating a conceptual tool for the analysis of other styles. As Wölfflin famously said in the preface to the first edition of Die klassische Kunst (1899),15 Hildebrand's book offered new techniques for the analysis of works of visual art taking it out of the domains of artistic personalities and historical circumstances. Wölfflin elaborated the contrast between optic and haptic into a set of five polarities which he then applied to Classic and Baroque art. He theorized the shift from optic to haptic, from one consistent set of formal categories to another set, as an autonomous development of art according to its own visual laws. Perhaps the most radical application of the new formalistic technique of analysis was 16 Riegl's. While there could be an argument for the autonomy of artistic development over a relatively short period of history, Riegl boldly applied the doctrine to a much greater period of time and wider sweep of cultures. The general perceptual thesis of Die spätrömische Kunstindustrie (1901)17 covered the development of the history of the visual arts from Egypt and archaic Greece, through classical Greece and Rome, to the early middle ages. All three modes of representation were to be taken as equally naturalistic, each reflecting the artist's engagement with the visual world: to reclaim the `naturalism' for particular styles can just lead to misunderstandings. The ancient Egyptian who tried to represent the objects in their strictly `objective' appearance meant this to be as `naturalistic' as one could imagine. The Greek, however, felt his own to be especially `naturalistic' when he compared his with them. And could the master of the portraits of Constantine with its lively expression of the eyes not have felt that he was a greater `naturalist' than, for example, the master of the portrait of Pericles? Yet all three would have, in the most modern sense, taken `naturalism' for something purely unnatural. Indeed, each style of art strives for a true representation of nature and nothing else and each has indeed its own perception of nature in that he views a very particular phenomenon of it (tactile or optical, Nahsicht, Normalsicht, or Fernsicht).18 Wölfflin had remarked on the asymmetry between visual attention and pictorial depiction: Hildebrand, p. 62. Hildebrand. p. 100. 15 Wölfflin, Classic Art, translated by P & L Murray (London, 1968). 16 For connections between Wölfflin and Riegl see Joan Hart, `Some Reflections on Wölfflin and the Vienna School', Akten des XXV Internationalen Kongresses für Kunstgeschichte (Vienna, 1984), Vol. I. pp. 53-64. 17 Alois Riegl, Die spätrömische Kunstindustrie nach den Funden in Österreich-Ungarn (Vienna, 1901) translated as Late Roman Art Industry (Rolf Winkes) (Rome, 1985). In view of this article's stress on historical continuities it should be said that although Riegl's Kunstindustrie was originally published in 1901, a second edition was published in 1927 in a collection of his writings, Gesammelte Aufsätze, edited by Hans Sedlmayr. Sedlmayr was a member of the group responsible for Kritische Berichte, where Gombrich made his first contributions to art historiography. Sedlmayr also published a book, Die Architektur Borrominis (Berlin, 1930), which set out to apply the techniques of Gestalt analysis to Borromini's architecture; the challenge which this posed to Gombrich was reported in his essay `Art History and Psychology in Vienna Fifty Years Ago', College Art Journal (Summer 1984), pp. 162-4. One finds echoes of the situation in Norm and Form (London, 1966), pp. 82-83. Gombrich's critique of Wölfflin did not appear until 1963 and was published in Norm and Form, under the title `Norm and Form'. 18 Riegl, p. 226, n. 117. 14 13 L. B. Alberti [1435] had observed that a person walking over a green meadow takes on a green colour in the face, but even to him this fact had no binding significance for painting. We see here how little style is determined by observations of nature alone, and that it is always decorative principles, convictions of taste, to which the last decision is assigned.19 By contrast, Riegl regarded works of art as products of will (Kunstwollen) which manifested the World View (Weltanschauung) of the age. With his hostility to considerations governing the technical development of art, the history of art became the history of forms of visual attention. Riegl, like Hildebrand, accepted the notion of the priority of the existence of a spatial field over the construction of forms within it. It was only within the pre-existent space of the visual field that one might perceive objects near to, middling or far away and those distances determined both the appearance of, and organizational relationships between, the objects depicted in the pictorial field. The three different stages of artistic construction could be equated to three different possibilities of subjecting the world to visual attention: (1) in the Egyptian and archaic Greek phases, there was an adherence to the `material individuality of objects ... the tactile plane suggested by the sense of touch'.20 From an optical point of view, the eye was so close to its subject that depthdefining shadows disappeared. `The main accent ... is placed on the silhouettes which are kept as symmetrical as possible because symmetry reveals to the exterior an uninterrupted tactile connection within the plane in the most convincing manner'.21(2) In classical Greece, the eye came into play recognizing the presence of depth through the existence of shadows. But at the same time the materiality of forms was retained through the perception of surface continuities. 'To perceive them [the forms] the eye has to move a little from the Nahsicht: not too far away, so that the uninterrupted tactile connection of the parts are no longer visible (Fernsicht) but rather to the middle ... we may call it Normalsicht . . .'. 22 (3) In the art of late antiquity `an existence of space appears to be recognised, but only as it adheres to material individuals; that is an impenetrable coherent space measured cubically, not infinite deep space between material objects'.23 Objects were located in relation to each other, not in space, but across the pictorial plane which was not, itself, tactile because it was broken by deep shadows. In summary, each phase in the development of depiction in antique art constituted an alternative mode of representing spatial entities generated by their presence as objects of near, medium and distant vision. In accordance with traditional associationist psychology, visible things were the product of the combination of sensations registered by the presence of forms and colours in planes and space.24 Riegl's analysis of the representation of space was taken further by Panofsky in his famous essay `Die Perspektive als symbolische Form' published in the Vorträge der Bibliothek Warburg for 1924-5.25 Although Panofsky started by describing `perspective in the full sense' as providing the spectator with an opportunity to look through the picture plane into a world circumscribed by the laws of one point linear perspective, he also made it clear that this was not the only valid spatial construction of the world. He contrasted mathematical/optical space with psycho-physiological 19 Wölfflin, Kunstgeschichteliche Grundbegriffe (1915), trans. M. D. Hottinger as Principles of Art History (London, 1932), p. 51. 20 Riegl, p. 24. 21 Riegl, p. 25. 22 Riegl, p. 25. 23 Riegl, p. 26. 24 Gombrich's expression, used in the Bodonyi Review; see note 33. 25 Now available in translation by Christopher S. Wood, Perspective as Symbolic Form/Erwin Panofsky (New York, 1991). This essay should properly be read in conjunction with `Die Entwicklung der Proportionslehre als Abbild der Stilentwicklung', Monatshefte für Kunstwissenschaft, XIV (1921), republished in translation in Meaning in the Visual Arts (New York. 1955). space, arguing that both were equally valid forms of representation, both articulated facets of experience that could be seen. It was not just that the artist adopted a convention of representation, rather that there was a clear symmetry between perception and depiction: Thus in an epoch whose perception was governed by a conception of space expressed by strict linear perspective, the curvatures of our, so to speak, spheroidal optical world had to be rediscovered. However, in a time that was accustomed to seeing perspectivally but not in linear perspective those curvatures were simply taken for granted: that is in antiquity. In antique optics and art theory (as well as in philosophy, although here only in the form of analogies) we constantly encounter the observation that straight lines are seen as curved and curved lines as straight;26 Echoing Riegl, he argued that in late antiquity there had been a major shift in the character of spatial perception. Classical antiquity had a systemic view of space, as a continuum in which objects were to be depicted and in terms of which they were to be seen in relation to each other. By contrast, as Late Antiquity did not know `systemraum' its artists proceeded from the concrete individual object, generating a 'Dingraum' which was an aggregate, or discontinuous, space. The history of art was related to a history of changing spatial perception, all of which was appropriate to the cultural moments in which it was produced. In a later essay 'Zum Problem der Beschreibung and Inhaltsdeutung von Werken der bildenden Kunst', published in 1932,27 Panofsky focused on the kind of difficulty that this could create for the art historian. If even a Roman, Lucian, looking at a painting by the Greek Zeuxis could not tell what the precise spatial relationships were between depicted objects,28 there must be a fundamental problem involved in the historian's activity of describing works of art. How does the art historian gain access to the contemporary perception of works of art produced in the past? According to Panofsky, it must be out of an intuitive sense of the validity of depictions born from cultural understandings of the workings of perceptual processes. It is important to stress that it would make no sense to ask whether, in Panofsky's view, optical perception changed through history. Given the necessary interaction between mind and eye, he held that the mind's construction of what was offered to it by the senses changed over time: the perception of the appearance of something can only acquire a linear or painterly form through the active intervention of the mind: so that the `optical focus' should be interpreted as a mental or spiritual focus on the optical, and consequently, the `connection of the eye to the world' is in truth a connection of the mind to the world of the eye. 29 Perception and depiction reflected each other and were the products of the cultural psychology of an epoch, of which any stylistic feature of an individual artist's work was a manifestation. Like any particular use of proportion or perspective, Michelangelo's form-defining hatchings, for example, were symptomatic of much deeper cultural and spiritual values. This approach to stylistic analysis tied together his early theoretical writings with the introduction to Studies in Iconology which was, itself, an extension of the essay on the problem of describing works of art. Panofsky, p. 34. This seems to me like a case of the El Greco fallacy but see Gombrich's rebuttal of this, which is to be found in Art and Illusion (Oxford, 1986), p. 218. 27 Panofsky, Logos, XXI (1932), pp. 103-119. 28 Panofsky referred to Lessing's Lettres Antiques for the source but see now the extract from Lucian's Zeuxis or Antiochos, 3, in J. J. Pollitt, The Art of Greece 1400-31 B.C. (Englewood Cliffs. 1965), pp. 156-158. 29 Quoted from `Das Problem des Stils in der bildenden Kunst', Zeitschrift far Ästhetik und allgemeine Kunstwissenschaft, 10 (1915), in Michael Ann Holly, Panofsky and the Foundations of Art History (New York, 1984). p. 61. 26 In the very year that Panofsky published his essay on the problem of description, Gombrich spent his summer thinking about the same problem, which was in the air'. His own meditations surfaced in the Bodonyi Review in 1935.30 Under the tuition of Julius von Schlosser, Gombrich had become deeply engaged with the notion of Kunstsprache, the visual mechanics of artistic imagery. Schlosser was a friend of Croce's and shared his view of art as language;31 he was also a friend of Karl Vossler and encouraged his students to study linguistics. Consequently, Gombrich attended the lectures of Karl Bühler who, besides holding the chair of psychology at the University of Vienna, was also deeply interested in linguistics. The Bodonyi Review was both an investigation of Kunstsprache and a re-examination of traditional psychological assumptions about the nature of pictorial imagery. One of the hallmarks of Bühler's approach was its combination of conceptual enquiry with experimental observation. The Bodonyi Review concerned itself with providing a testable framework for the interpretation of data drawn from history.32 Its subject was the use of the gold ground in late antique art and the old explanatory theory held up for re-examination was Riegl's formalist conception of the work of art as only form and colour in plane and space.33 Sharing Schlosser's hostility to the thematic study of motifs, it treated the gold ground as a communicative device. The first stage of the analysis explicitly foregrounded psychological explanation. It drew from Bühler's study Die Erscheinungsweisen der Farben,34 to characterize the gold as a compacted surface, like fallen snow. Not really perceived as the boundary to a space, defining space, it appeared to swim beyond it. As a surface it offered itself to the artist who wants to represent light. But gold did not just represent light, it was also a symbol of light35 and the symbolic meaning of light had been spelt out by the church fathers in their commentaries on biblical passages such as `For with thee is the fountain of life: in thy light shall we see light' in the Old Testament and `God is light' in the New.36 Furthermore, gold was not one colour among many but a metal with a material presence. The gold ground broke the illusion of art to present the spectator with a materially valuable thing, the type of object of which the primitive bejewelled icon was the best example. Following analysis of the perceptual effects of the gold ground, the next stage was to consider its role in the total image, much as one might try to analyse a text within a larger and more complex text.37 Given the possibility that the gold ground represented light, the question emerged as `J. Bodonyi, Entstehung und Bedeutung des Goldgrundes in der spätantiken Bildkomposition (Archaeologiai Ertesitë, 46, 79.32/3)'. Kritische Berichte zür Kunstgeschichtlichen Literatur, 5, 1932/33 (published in 1935), pp. 65-75. At the beginning of the review, Gombrich mentioned an exchange of views which occurred between himself and Bodonyi at the time of the work's composition; the review was a continuation of that discus-sion, not a `critique or criticism'. Irrespect ive of the contents of Bodonyi's dissertation, the contents of the Review were Gombrich's own. 31 On Schlosser and language see Michael Podro, `Against Formalism: Schlosser on Stilgeschichte', Akten des XXV Interna tionalen Kongresses für Kunstgeschichte (Vienna, 1984), Vol. I, pp. 37-43. 32 Another historical experiment that Gombrich became involved in before he started working as Ernst Kris's assistant was E. Kris and O. Kurz, Die Legende vom Kunstler: Ein historischer Versuch (Vienna, 1934). 33 `Die neue Grundauffassung wird schon im Untertitel der Arbeit angekündigt: "Ein Beitrag zur Sinndeutung der spätantiken Kunstsprache." Damit ist die Stellung zu Riegls berühmter These gekennzeichnet, die im Kunstwerk nu "Form und Farbe in Ebene und Raum" sehen will. Es ist die damals herrschende Formalästhetik Herbarts, ihm wohl durch Zimmermann in Wien vermittelt, aus der sich Riegl in seine m ersten Hauptwerk das methodische Rüstzeug zu schaffen sucht.' Gombrich, Bodonyi Review, p. 66. 34 Bühler, Die Erscheinungsweisten der Farben (Jena, 1922). 35 Gombrich raised the question of why if it was available to represent light, it hadn't been used by naturalistic artists? He gave a full answer to this question in `Visual Meta phors of Value in Art' (1952), republished in Meditations on a Hobby-Horse (London, 1963). 36 Psalm 36, 9 & John Epistle I. 1. 5. 37 'Er wählt also seinen Gegenstand, nicht weil er ihn durchaus für das älteste Beispiel für das Auftreten des Goldgrundes hält; seine Gründe sind vielmehr methodischer und zwar wieder sprachgeschichtlicher Natur: So wie ein unbekanntes Wort und seine Bedeutung leichter in einem größeren Textzusammenhang ermittelt 30 to how it fitted into the composition to form a coherent whole. Its pictorial presence had to be given sense; the notion of compositional effect required an understanding of what was `going on'.38 Formal analysis had to give way to a third strategy, that of genetic analysis: the historical investigation of the development of a particular pictorial construction.39 Acknowledging Panofsky's study of the development of the representation of space, Gombrich criticized his notion of Dingraum' aggregate or discontinuous space describing it as the hypostatization of a negative. `Lack of unity is not a special form of unity'.40 In the Albani landscape, the separate landscape zones were unified by light-flooded space. As a pictorial technique, this kind of unification could also be seen in East Asian landscapes of cloud and mist. In the case of the mosaics at S. Maria Maggiore, the gold ground occupied the space between foreground and background but commentators made the mistake of treating it naturalistically: Pfuhl as the representation of a flood, Sieveking as a rocky path and Kömstedt, in yet another example, as a sandy beach. Close inspection for consistency revealed the invalidity of such interpretations. Genetic analysis demonstrated that the gold ground stood between the foreground and background as a neutral, light-emanating, space; it both represented light and was light.41 The development of the neutral space of the gold ground emerged out of the collapse of representational techniques used to depict both space and colour. The transformation of the nature of the middle ground demanded historical explanation. Riegl answered in terms of a changed Kunstwollen from haptic to optic perception. Paradoxically, this preserved a naturalist approach to the representation of space: the picture plane was still treated as being transparent onto a world. Riegl treated changes in the history of the human spiritual condition as an important part of his explanation but simply as a parallel to the artistic changes. Gombrich criticized this as failing to provide a causal explanation of the change: `one has to think of a kind of prestabilised harmony of autonomous streams, not of the interactions of cogs in a machine'42 Gombrich's criticism of Dvorak could equally be applied to Riegl: `The unity of the art concept (Kunstbegriff), of the spiritual and social function of art within different spiritual/intellectual werden kann, so ist für B. das Problem des Goldgrundes dort am ehesten zu lösen, wo es am kom plexesten erscheint, d.h. wo der Goldgrund "in eigenartiger Weise in die Komposition einbezogen ist".' Gombrich, Bodonyi Review, pp. 67-8. For a demonstration of this point see Gombrich's remarks on Veronese's Mystical Marriage of St Catherine in `The Meaning of Beauty' reprinted in E. H. Gombrich, Reflections on the History of Art, ed. Richard Woodfield (Oxford, 1987), p. 188. 39 39 Genetic analysis has been a constant feature of Gombrich's work. Art and Illusion is, of course, the major study but other examples are `The Heritage of Apelles' in The Heritage of Apelles (Oxford, 1976), `Ideal and Type in Italian Renaissance Painting' in New Light on Old Masters (Oxford, 1986) and, more recently, 'Watching Artists at Work: Commitment and Improvisation in the History of Drawing' in Topics of our Time (London, 1991). 40 `Panofsky spricht vom Dingraum, Aggrega traum, vom diskontinuierlichen Raum. Es scheint mir wichtig auf die Gefahr hin zuweisen, die jeder Hypostasierung solcher Negativa droht: Uneinheitlichkeit ist keine Spezialform der Einheit, Aggregatraum auch keine des Raumes im neuzeitlichen Sinne.' Gombrich, Bodonyi Review, p. 68. 41 As a proof by negative example, Gombrich cited a scene in the depiction of the Mount of Olives in the Codex Rosanensis where there was a strip of darkness as a sign of night between the moonlit earth and the zone of the starry sky. 42 'Die tiefe geistesgeschichtliche Wandlung wird bei Riegl natürlich nicht außer Acht gelassen. Sie wird aber der Änderung des Gestaltungswillens gleichsam koordiniert als parallelgerichtetes Kulturwollen etwa, das vergleichbare, strukturverwandte Erscheinungen aufweise. Man hat wohl an ein Art prästabilierter Harmonie autonomer Abläufe zu denken, nicht an das Ineinandergreifen eines Räderwerks.' Gombrich, Bodonyi Review, p. 71. 38 circles and constellations, is unreflexively assumed'.43 He condemned it as nothing less than a thoroughly unhistorical aestheticism. Reinstating history required a sociological investigation of the changed cultural function of art in late antiquity. For Gombrich, perception of form could not be divorced from perception of meaning. Recognizing that pictures mediated experience of their created world, he emphasized the necessity of understanding the nature of that mediation and the function of imagery at the time that it was produced: form followed function. Riegl had simply failed to recognize the changed representational function of pictorial art in late antiquity. In the art of classical antiquity, imagery was created in which the viewer was expected to assume the role of a spectator in a scene in which he had an imagined presence. The depiction of the scene was governed by the notion of the fruitful moment, which was not the climax but the turning point of action. The turning point was particularly effective because it would generate a variety of imaginative responses to the depicted event44 By contrast, in the art of late antiquity the purpose of visual imagery was quite different as the visual arts had changed their place within the overall cultural framework; their role had changed from `showing' to `telling'.45 This, in turn, brought with it different formal structures. The dematerialization of the figure, the replacement of mimicry by set expressive gestures and the significance of gesture overall were structural properties which belonged to the picture-script (Bilderschrift). The idea of the picture script was signalled in the famous letter from St Gregory the Great to Bishop Serenus of Marseille: `What Scripture is to the educated, images are to the ignorant, who see through them what they must accept; they read in them what they cannot read in books'.46 The compositional difference between the two visual forms, the scene and the picture- script, lay in their principles of unity. In the later art: `All parts of the picture are virtually at the same time the words of a sentence. Only their summary gives a meaning: the content of the story.47 In terms of pictorial logic, not only was the idea of the fruitful moment inapplicable to picture-script but a picture field containing continuous narrative was as little a slice of space as it was a certain moment. Picture-scripts consisted in juxtapositions of motifs and if motifs were juxtaposed they could not be understood in relations of existence to each other in a possible depicted space: 'Discontinuous space is not really a space at all but involves a quite different representational principle.'48 To take another, and more obvious, example, if, in the case of continuous narrative, the same figure is repeated across the pictorial field it cannot be said to be to the left or the right of itself. But there was another important factor to take into account and that was the difference between the abbreviated signs of the catacombs and the picture scripts of the mosaics. The mosaics were symbols in Goethe's sense, symbols as opposed to rationally constructed signs. The symbol was not exhausted by meaning but was a sensuous manifestation of it; in a magical sense, it was what it represented.49 The question of the aesthetic value of the art of late antiquity should not be thought of in terms of the formal values of a world of illusion (Scheinwelt) but in terms of the materiality of the Die Einheit des Kunstbegriffs, der seelischen und sozialen Funktion der Kunst innerhalb verschiedener geistiger und kultureller Kreise und Konstellationen wird unbedenklich vorausgesetzt ...' Gombrich, Bodonyi Review, p. 72. 44 `Man könnte sagen, nicht die Aktion. die Reaktion stand im Mittelpunkt.' Ibid.. p. 72. 45 The contrast between showing and telling was brought out at length in chapter 6 of The Story of Art. 46 Translated in C. Davis-Weyer, Early Medieval Art 300-1150 (Englewood Cliffs, 1971), PP. 47-49. 47 Gombrich quoting Bodonyi: `Alle Teile des Bildes sind gleichsam die Worte eines Satzes. Erst ihre Zusammenfassung gibt einen Sinn: den Inhalt der Erzählung.' Gombrich, Bodonyi Review, p. 73. 48 Wird besonders klar, daß der diskontinuierliche Raum eben gar kein Raum ist, sondern ein ganz anderes Darstellungsprinzip involviert.' Gombrich. Bodonyi Review, p. 71. Fo r recent observations of this kind see John Hyman, The Imitation of Nature (Cambridge, 1989), pp. 103 ff. 49 It was precisely within this context that one should understand the Iconoclastic Controversy. 43 artefact.50 This was an issue which Gombrich took further in his second contribution to Kritische Berichte, the Garger Review of 1937.51 The fundamental mistake made by the formalist writers following Fiedler was to assume that there was a straightforward relationship between the artist's picture plane and the so-called visual field. Extreme formalism was vitiated by the need to identify coloured forms as objects and expressively active people as expressively active.52 The conclusive objection to formalism was, however, that one needed to understand pictorial conventions to describe relationships in the pictorial field. This was not so much a matter of Stilgeschichte as Sprachgeschichte: Kunstsprache rather than style, device rather than motif, artefact rather than work of art.53 The starting point for understanding such conventions was the institutional context of artistic practice, the overall purpose behind artistic production within a specific culture. As Gombrich and Kris declared in their later unpublished manuscript on caricature: `there is no unambiguous "language of form"such as artists and aestheticians have sometimes dreamed of. Any such language only comes into being in an institutional context. 54 Paradoxically, Hildebrandian formalism could not account for the value of the artifactuality of the work of art and, in so far as the aesthetic quality of late antique art depended on colour, neither could Fry's formalism. 51 'Wertprobleme und mittelalterliche Kunst', Kritische Berichte, 1937, translated and published in Meditations on a Hobby-Horse, London, 1963, as 'Achievement in Medieval Art'. Panofsky discussed this at length in his 1932 essay on the problem of description. Pespective as Symbolic Form/Erwin Panofsky. p. 237. The following is the 1939 shortened version in Studies in Iconology: ' "Formal analysis" in Wölfflin's sense is largely an analysis of motifs and combinations of motifs (compositions): for a formal analysis in the strict sense of the word would e ven have to avoid such expressions as "man", "horse", or "column", let alone such evaluations as "the ugly triangle between the legs of Michelangelo's David" or "the admirable clarification of the joints in a human body".' (p. 7). 53 On which, for Schlosser's point of view see Julius von Schlosser, ' "Stilgeschichte" und "Sprachgeschichte" der bildenden Kunst', Sitzungberichte der Bayerischen Akademie der Wissenschaften, Phil. Hist. Abt. 1 (1935), pp. 3-39. 54 Unpublished MS (1938), The Caricature Style', p. 3. 52 50