The cleaning of the Sistine ceiling in the 1980s brought unexpected insights into Michelangelo’s approach to painting. The dirty surface had contributed to the idea that the artist was not much interested in colour and was more obsessed with forms and especially the form of the human nude. Nevertheless, the cleaning revealed what appeared to be unusual colour combinations. In addition to modelling with single hues, from light to saturated, Michelangelo also used different colours for shadows and highlights. The lunettes, however, presented interesting challenges. There, one found especially contrasting colours. The Asa-Josaphat-Joram lunette, for example, features the Asa figure on the left with a cloak modelled from apple green to deep orange and red.

In the symposium dedicated to the unveiling of the restored frescoes, costume historian Edward Maeder interpreted these colours in the lunettes as ‘shot-silk’ – in Italian, cangianti, a kind of silk woven with contrasting colours that appear iridescent when seen in the light. According to Maeder, who had singled out the figures of the ancestors in his arguments, colour changes are used to make these Eastern ancestors of Christ more exotic. While it is true that colour changes have been used to represent shot-silk in Italian painting, Maeder went on to suggest that since this fabric was most commonly available from the Near East, it would carry Oriental associations and therefore serve ideally for the ancestors.

Maeder constructed his argument in a typical art historical way: we see figures that evoke associative meanings. But he was challenged strongly by the English art historian, John Shearman. Shearman had written a dissertation on the colour of Tuscan painters like Michelangelo and emphasised the great artist’s dependence on his teacher, Ghirlandaio, an artist who had practised the Tuscan tradition of ‘isochromatic colour composition’ in the ceiling. This system, according to which broad, flat fields of colour are decoratively balanced along the picture surface, is frankly medieval in origin. According to Shearman, decorative fresco was required to create surface unities to aid visibility even more than making possible narrative associations.

Years later, Shearman prominently titled his contribution to the postrestoration volume on the Sistine ceiling, ‘The Functions of Michelangelo’s Colour’, explaining how the high-key colours were in large part a response to the conditions of visibility, illumination and decoration that Michelangelo was facing.

The work that announced the methodology of the New Vienna School was Hans Sedlmayr’s Die Architektur Borrominis of 1930. A brilliant and difficult book, it has had scattered influence but, like all of Sedlmayr’s work, is tainted by its author’s name. Martin Raspe, who attempted a similar syncretic view of the architect to move beyond the standard catalogue was at pains to distance himself from Sedlmayr’s example, which loomed so large in his own German-language tradition. Indeed, language has been a barrier to fairly assessing Sedlmayr, as was his own notorious characterological interpretation of Borromini. But the book advances a remarkable analysis of Borromini’s work, which is worth the effort to understand.

The model of Sedlmayr’s self-reflexivity that I have sketched in the last two chapters can be confirmed in the book on Borromini. Indeed, if the author gets ahead of himself with his later speculations on Borromini’s personality, the overall aim of the book is reasonable – to give a phenomenological account of Borromini’s architecture. This is the propadeutic use of phenomenology as a basis for any critical discipline, in line with the appropriation of Kurt Lewin’s idea of a proper vergleichende Wissenschaftstheorie. If Sedlmayr’s approach is self-conscious about its own contribution, then we have to regard it not only according to its stated aim but the discipline of architectural history at the time. With his lens of Gestalt psychology, Sedlmayr achieves a remarkable deal of self-reflexivity in his discussion of Borromini.

After expanding the idea of Sedlmayr’s method, I will pass on to a full exposition of his analysis. Next, I undertake a full discussion of Sedlmayr’s notorious use of characterology – the use of the morphology of the body to infer into a person’s personality – as a means to make conclusions about Borromini’s own personality. While I fault Sedlmayr’s reliance on Ernst Kretschmer’s ideas, I also clarify how many things thought to be true about the theory are in fact inaccurate.

If Hans Sedlmayr’s monograph on Borromini stated the outlines of the method of Strukturforschung, its next most impressive articulation came in Otto Pächt’s extensive discussion of the formative principles found in Late Gothic painting in Northern Europe in his ‘Design Principles of Fifteenth- Century Northern Painting’, published in the second and last volume of Kunstwissenschaftliche Forschungen. This lengthy study is just as influential as Sedlmayr’s, and indeed more rigorous and imitable in many ways. It is furthermore methodologically self-aware and usefully stakes out a scholarly way of proceeding that is aware of its competitors, namely Panofsky’s Iconology.

The most interesting and controversial element developed by Pächt was the fruitfulness of using concepts of ‘national’ constants in investigating art history. Pächt argued that in examining the variety of Late Medieval painting in western Europe, he could define distinct ‘constants’ that involved the relationship of figural content to ground. For example, he found that Netherlandish (not Flemish) is characterised by a flat surface, in which figure and ground interlock on the surface. As we shall see, in understanding such a concept everything hangs on what ‘national’ means.

Pächt utilised this theory throughout his long career and it found echoes in his students’ work. Like Sedlmayr’s characterological analysis of Borromini, his method has a speculative and potentially suspicious element attached to it. Pächt was criticised quickly by Meyer Schapiro and while many agree that Pächt practised the method scrupulously, its findings were immediately tarnished (as Pächt himself saw) – a confirmation of the instability of such concepts for those more critical like Schapiro – by the recklessness of those engaged in Kunstgeographie and especially Ostforschung.

Always less objectionable in studies of anonymous ancient art (in the same volume of Kunstwissenschaftliche Forschungen, Kaschnitz-Weinberg’s paper on Egyptian art appeared), the idea of stylistic constants continues to have its adherents, sometime inflected with semiotic theory (which perhaps assures the reader it is entirely ‘conventional’), but for the most part is considered a dead proposition.

If Strukturforschung was devoted to the integrity of the work of art it also had a theory of history that, in a sense, was committed to the integrity of history. Jan Bakoš calls this a ‘structural evolutionary conception of art history’. One could say that in this conception the structure of the work of interlocking that moved real history beyond mere ‘genetic’ links of the ‘first’ task of art history. This evolutionism ultimately comes to terms with Riegl’s mysterious notion of the Kunstwollen, which both Sedlmayr and Pächt embrace with qualifications.

The argument – made on empirical grounds – is that contemporary theory has validated Riegl. The result of course was an apparent holistic form of determinism, with obvious ominous tones when seen in retrospect.

Of course, both Riegl’s Kunstwollen and the Vienna approach to Strukturforschung hit a major brick wall in Gombrich’s influential critique of both in Art and Illusion.

It is difficult to find another historical study of the Vienna group of the calibre of ‘Borromini’s architecture’ or ‘the design principles of Late Gothic painting’. To be sure, Bruno Fürst, who we saw to be a close friend of Pächt, produced a study of Austrian sculpture that in many ways extends Pächt’s categories of Gestaltungsprinzipien to three dimensions. Swoboda, so close to the group, did not produce anything characteristically structure-oriented apart from works continuing Pächt’s ideas on national constants. I will argue that Swoboda’s co-editor of Dvořák’s work, Johannes Wilde (1891–1970), deserves this attention. As we will see in a moment, Wilde – who was born in Budapest but took his PhD in Vienna with Dvořák – was considered integral to the University Strukturforschung group from his perch at the Gemäldegalerie of the Kunsthistorisches Museum of Vienna.

Instead of exploring studies produced by Wilde in Vienna in the late 1920s or early 1930s, however, I will focus – paralleling my previous discussion of Pächt’s works – on his work done in England. I will argue that Wilde used his English experience to filter his initially quite rich and varied philosophical background – derived from Dvořák and based loosely on what can be called ‘hermeneutic’ philosophy – to obtain a simplified functional approach to interpretation. This method has a lot in common with the method of Sedlmayr and Pächt, but with only direct influence and more of a common evolution, a homologous method you might say.

Wilde, like Sedlmayr and Pächt, was extremely sensitive to artistic form and its organisation. Whereas Sedlmayr had analysed the internal structure of Borromini’s San Carlo, Wilde was interested in the way in which frescoes were related to the space external to them. In this his interests might be likened to those of Pächt exploring the relationship of the miniature to the manuscript page.

Art history is awash in ‘theory’ but has no matrix to debate questions. At a moment when art history has searched for a healthier model of balanced theoretical reflection and empirical practice, emblematised by the recent translation of Erwin Panofsky’s early works, the second Vienna School emerges as an exemplar with impressive theoretical writings and incisive historical studies. This book has sought to contextualise this school and connect its theoretical commitments to currents of Gestalt theory, in order to better understand its scope and ambitions. Brought under the same umbrella, the New Vienna School is engaged in a dual challenge for the human sciences: to combine natural scientific rigour with hermeneutic understanding.

The New Vienna School exercised what I have called ‘analytic holism’, a naturalistic view of the interaction of forms – parts, wholes – in a relational manner amenable to analysis. By transferring the work of art history to the visual, the school emphasised the refinement of tools related to visual analysis. Gestalt theory showed how those tools could be rigorous. Indeed, by being thoroughly sensory, they were the prerequisite for any formalisation of a theory of symbolism or style in art history.

This book has tried to stress the reflexive awareness of the tasks of art history by the new Viennese group, a reflexive awareness that it recognised in, and adopted from, Gestalt theory. The balance of theory and practice in the writings of Hans Sedlmayr and Otto Pächt is not accidental, as each saw the positioning of the discipline to be just as important as the work that could be undertaken within it. As I explained, they understood that there were many problematic notions facing a rigorous art history, beginning with the meaning of individual forms and their possible connections. Style history and iconography presented two interpretive practices that ran into methodological difficulties; namely, what parts or works can be compared fruitfully, and what parts or works can we attach ideas to?

The final chapter on Otto Demus moved problems of visuality beyond the Latin west, and indeed there are many more directions in which Viennese theory can lead us.

The main task that challenged Sedlmayr and Pächt in the late 1920s to early 1930s was acknowledging the integrity of the work of art. From Gestalt theory they inherited the concern that one must first understand the basic structure and nature of a work of art before attempting to understand its history. They saw many of their contemporaries as failing in this task and therefore incapable of making meaningful progress in art history. The discipline was full of tacit assumptions about the work of art, the nature of representation and symbolisation, and historical influence that needed to be made explicit and openly theorised. The place to begin was with the single work itself.

In his 1930 book on Borromini, Sedlmayr argued that one would have to have an understanding of the basic structural principles constituting the architect’s style before understanding the Baroque. It was literally impossible, in Sedlmayr’s view, for one to understand the genealogy of Baroque architecture without first having an explicit account of Borromini. Similarly, when Pächt addressed the problem of regional styles in Netherlandish painting in his 1933 study of ‘Design Principles’, he argued that it is hopeless to argue which artists were influenced by which traditions without articulating accounts of just what those regional styles entailed, their inner form or Gestalungsprinzip.

Sedlmayr and Pächt felt that the tools to move beyond an informal formalism could be found in Gestalt psychology, which offered the means to characterise the structure of visual perception that could be applied to individual works of art and architecture. Gestalt psychology had even broached the problem of development and identity of structure in time. The theory developed by the early Sedlmayr and mature Pächt presents a compact set of commitments, stressing spatial complexes in relation to viewers, the causes of which are contiguous, immediately antecedent events. With the view that the New Vienna School is devoted to a kind of ‘structural evolutionism’, I will devote this chapter to structure and the next to evolution.

When attention began to turn towards the second or New Vienna School, scholars grouped researchers together in an inclusive approach. Thus, Fritz Novotny’s work on Cezanne was included in Christopher Wood’s Vienna School Reader, even though Novotny had been a product of Strzygowski’s Institute and not (the descendent of) Riegl’s. Subsequently, his membership in the second Vienna School has been questioned. The same issue faces another student of Stzrygowsky: Otto Demus. Nevertheless, in the following I will emphasise Demus’s academic socialisation after completing his university training, whereupon he actually did teach at the University of Vienna with Pächt. By the 1960s, the ‘two Ottos’ of course would become famous for having offered one of the most formidable medieval art programmes in the world.

As I will argue, Demus’s work fits quite well with the second Viennese school to the degree that his work is focused on an understanding of the work of art or monument as a functional whole. Parts and their relationship are understood, or in the case of a lost work, intuited by way of reconstruction. In addition, technical knowledge is used in order to judge initial states of objects so that the universal working of perception will be accurate based on these same givens.

I propose to explain Demus’s methodology through his most famous contribution: an exegesis of the structure and function of Byzantine mosaics. I will show that his approach, which has been linked to Riegl in a casual way, can be aligned in a more rigorous way. After reviewing Demus’s career, I will pass on to his important interpretation of Byzantine mosaic decoration in terms of negative perspective and the icon in space. These considerations challenge traditional notions of the self-sufficient image and mark the passage of the image as space. If Wilde had placed the image into its relational context, Demus instead marked the movement of figure and ground into new spatial configurations.

In 1925, a young, ambitious art historian, Hans Sedlmayr, published an article on Francesco Borromini called ‘Formed Seeing’ (‘Gestaltetes Sehen’). The article could be called a manifesto, as it self-consciously declared a new theoretical and methodological point of view. Against previous attempts to understand Francesco Borromini based on theological and geometric premises, Sedlmayr wishes to press the question of rigour (die Strenge) and challenge art historians to develop interpretations that were founded in the visual material itself. Through the understanding of artistic organisation, inspired by the then emerging findings of Gestalt psychology, the art historian could go beyond intuitionist style history or iconographic symbol hunting. Sedlmayr was joined immediately by a small group of scholars – in particular Otto Pächt – and they anxiously promoted their ‘new’ Vienna School with Riegl as their idol.

The new platform of Strukturforschung was born in a self-conscious understanding of intellectual movements in the larger sciences and had a clear idea of how one discipline had addressed the ongoing ‘crisis of the sciences’ (Krisis der Wissenschaften): Gestalt theory. With the disruption of the First World War and social upheaval in both the newly founded Weimar Germany and the First Austrian Republic, there was widespread fear that the riches of the modern technological state were not in alignment with the traditional needs of social life. The meaning of modern science was everywhere debated, and many sought a reconciliation between subjective knowledge and natural science.

Gestalt theory began to be articulated in the second decade of the twentieth century and provided one important response to this crisis. The solution was a monistic approach to mental life that, in the traditional of Naturalphilosophie, sought the forms of thoughts in natural configurations. The original Gestaltists – Max Wertheimer, Wolfgang Köhler and Kurt Koffka – balanced life on the razor’s edge of a phenomenal and scientific realism. Almost uniquely among Weimar theories, Gestalt theory refused to confirm its place either on the side of experience or science.