The Hans Sedlmayr—Rudolf Wittkower Debate (1930–1932)

Translated by Daniel Spaulding

1. Hans Sedlmayr, “Eduard Coudenhove-Erthal, Carlo Fontana und die Architektur des Römischen Spätbarocks. – Wien, Schroll. 1930.”

Book review, Kritische Berichte 3/4 (1930/31 and 1931/32), 93–95.

It has not been sufficiently understood, or sufficiently kept in mind, that wherever art historians today pursue their scholarly work two entirely distinct sciences of art arise side by side. If it were to be completed, the first would contain every piece of knowledge that can be obtained without understanding artworks as artworks: for example, assigning sequences of dates, reconstructing the external inventory of artworks, comparing works from extra-artistic viewpoints, etc. The other science of art, the second, searches out those items of knowledge that are the prerequisite for an understanding of artistic form as such: and it seeks this understanding itself. This second science of art is still in its infancy, whereas the first has already achieved definitive results; only the two together would constitute a true and complete science of art.—Recognition of this state of affairs is the precondition for a fair judgment of what has been in achieved in any individual case.

The book in question—well-equipped by the publisher—is a fine achievement of the first science of art and is in this respect a definitive work. It will be possible to supplement this volume—for example, with the drawings by Fontana that Wittkower has discovered in London; it is a characteristic of works of the first science of art that they can be added to like albums of stamps—but it renders a second book of this kind on Fontana absolutely superfluous. However, like Hempel’s monograph on Borromini, for example, this work, too, longs to be supplemented with a monograph belonging to the second science of art on the same (external) theme. Only the two together would contain everything that can presently be known of Fontana.

Monographs of this second kind must always be written anew; it is thus fortunate for Fontana that there is still no theory of his art. The sketches of Fontana’s art in Gurlitt, Escher (Thieme-Becker’s Künstlerlexikon XII, 1916), and now in Coudenhove are so little concrete and so pallid that one need not even bother to “erase” them. The following suggestions are meant to prepare observational material for a future monograph on the art of Fontana.

So long as one examines Fontana’s works sequentially, without having understood their formative principles, it would hardly seem credible that this sometimes oppressively dreary academic opportunist could have had ideas that are among the most modern and essential between Borromini and Fischer von Erlach. One would then probably take San Marcello as his major work—the most agreeable rendition of Bernini’s widely circulated concepts. The true major work of his early period (at least among those known to us) is, however, San Biagio in Campitelli.1 What is decisive for an understanding of this work is to see that the upper story is another spatial unfolding of the lower story and that, for a more distant beholder, this upper story merges into a repetition of the lower, as we can see quite nicely in an elevation from Rossi’s Studio.2 Above the invariant lower story, the upper story has two existences, so to speak: a “near view,” in which the upper story has almost nothing immediately to do with the theme of the lower story, and a “distant view,” in which the upper story repeats the theme of the lower story and is integrated with the latter into a single plane. (It cannot be objected that the distant view would have been physically impossible in the narrow alley, for in order to gain this view it is not unavoidably necessary to turn oneself around bodily. When viewed correctly, one experiences a curiously unreal turning-away from the building around the plane of its axis, and at the same time a fusion of the building with the surrounding space that functionally belongs to it. The view from the corner is possible, but this is a mere first impression, not the final goal of the composition.) Here the upper half of the form shifts—and thus the whole shifts, too. This is a rather ingenious formulation of the late baroque principle of “shifting vision” [verwandelndes Sehen], and also a very early one: 1664. It ought to be compared in detail with the more or less contemporaneous but typically high baroque approach to a similar idea in the façade of San Carlino.—It is accordingly no longer so difficult to understand why it is that Fontana made such a strong impression on the Austrians (Hildebrant, but also Fischer: Johanisspitalskirche!).3 The latter make a positive synthesis out of what in Fontana often only seems to be compromise.—The plan for a church in the Colosseum has a similar importance for Fontana’s late period (end of the 1690s). What is important here is not the content of the invention, but rather the extremely original development of a typically late baroque principle of design, without an understanding of which the idea would necessarily appear monstrous. In the lower zone, the body of the church and the oval arcade are drawn into a single form that one perceives as a forceful plastic-corporeal loop with a convex bulge in the center. In the second story, however, the same primary structure, flanked by the frontal wings of the towers and standing out against the backdrop of the Colosseum’s receding, evanescent mountain of ruins, strikes the viewer as something like a concave landscape panorama with an architectonic “foreground” in the center. The shading in the print makes this intention quite clear, although it hardly could have been carried out in reality. (We perhaps understand here better than anywhere else the significance that the ruinous can have for the late baroque: dissolution of a plastic aggregate state into the optical.) This entanglement of aspects adds a very novel—and very Italian—variant to the list of incarnations of the central idea of the Italian baroque.

Fontana’s art has a methodological interest for the art historian, in addition to a “material” one: it is a particularly convenient starting point for remarking the objective characteristics of an immediately perceived “quality.” It ought not to be difficult to agree that the façade of San Biagio makes a weak impression, an impression of “second” (if not third) quality. As an aesthetic event it compares poorly to the façade of San Carlino, to say nothing of Michelangelo’s façade of San Lorenzo, if we were to imagine the latter carried out. And this impression persists as a primary, inalterable factum even after one is convinced that the concept of the façade is by no means flat but rather quite fine and original (even if ultimately lacking in substance and depth). This is due—as one could experimentally demonstrate—to the inadequacy of the design in its smallest details. The idea of the whole would have demanded different moldings and detail motifs, which one practically could derive from a careful apprehension of the larger motif.

The lesson to be drawn from this case for a still urgently needed doctrine of the value of artistic forms is the observation that a discrepancy between the design as a whole and its details is perceptible even before one has adequately comprehended the design’s meaning. This seems quite plausible, even if one theoretically examines the facts of the matter on the basis of more general experience.

Vienna, December 1930

  • 1 [Translator’s note: This church, now deconsecrated, was in 1900 rededicated as Santa Rita da Cascia in Campitelli. Originally located near the foot of the stairs leading to Santa Maria Aracoeli, it was dismantled in 1928 as a result of the Fascist renovation of the area around the Capitoline Hill. In 1940 it was rebuilt at a new location perhaps a hundred yards away, closer to the Theater of Marcellus. The sightlines that are key to both Sedlmayr’s and Wittkower’s analyses of the building thus no longer exist. At the time these texts were published, the church indeed only “existed” as elements in storage, and of course as represented in earlier drawings, prints, and photographs. Although Sedlmayr does not discuss the dismantling of the church at all, while Wittkower mentions it only in passing, these events perhaps cast an interesting light on the virtual or imputed character of Sedlmayr’s non-empirical (indeed, impossible) “correct” view of the façade.]
  • 2 Domenico de’ Rossi, Studio d’architettura civile (Augsburg: Johann Ulrich Kraussen, 1716).
  • 3 [Translator’s note: The Johansspitalkirche, or St. Johanns-Kirche, in Salzburg, built 1695–1703.]
View of the Capitoline Hill, with the façade of San Biagio in Campitelli (left), Rome, 1665, designed by Carlo Fontana. Engraving by Giovanni Battista Falda in Giovanni Giacomo de Rossi, Il nuovo teatro delle fabriche, et edificii, in prospettiva di Roma moderna (Rome, 1665).

2. Rudolf Wittkower, “On Hans Sedlmayr’s Review of: E. Coudenhove-Erthal, Carlo Fontana

Originally published as “Zu Hans Sedlmayrs Besprechung von E. Coudenhove-Erthal: Carlo Fontana,” Kritische Berichte 3/4 (1930/31 and 1931/32), 142–145.

It is not usual to polemicize against book reviews. If I here respond to Sedlmayr’s remarks on a study of Carlo Fontana in the previous issue of this journal, it is for very particular reasons. Sedlmayr abstains from a critique of the Fontana book. Instead he offers his own contribution to the topic.

Since it is here a matter of criticizing criticism, I will not attempt to complete Sedlmayr’s actual task—to review Coudenhove-Erthal’s book.1

We also learn the reasons why Sedlmayr has neglected to write a “critical report” [kritischer Bericht]. In general, Sedlmayr claims that Coudenhove-Erthal’s book is a fine collection of materials, though assembled according to a non-artistic point of view: a “definitive work” of “first art history” (which, as we have already seen, is precisely not the case). Sedlmayr sharply distinguishes this from those approaches that address artworks themselves, with their structures and their hidden laws. (The “second art history.” Cf. Sedlmayr in Kunstwissenschaftliche Forschungen, issue 1.)2 Sedlmayr considers it his task to evaluate the material that Coudenhove-Erthal has provided in terms of this “second art history.” Along these lines, he offers “suggestions [that] are meant to prepare observational material for a future monograph on the art of Fontana.” Sedlmayr thus raises the reader’s expectation that his own remarks will differ in principle from Coudenhove-Erthal’s interpretations; they will instead belong to a higher level of observation, so to speak. Against this, we can affirm that Coudenhove-Erthal honestly attempts a characterization of Fontana’s art. Even if these efforts meet with doubtful success and are pursued along lines that Sedlmayr and others consider faulty, this in no way demonstrates that the starting points and aims of the investigation are different: in fact, they are the same. In apparent loyalty to the reviewer’s duty to offer opinion and correction, Sedlmayr in truth divests himself of it.

After the claims with which Sedlmayr introduces his own suggestions, we might expect at least to find our understanding of Carlo Fontana substantially enriched. The result is disappointing. In two of the three cases he examines, Sedlmayr falls behind Coudenhove-Erthal: not, indeed, in the specificity of his diction but rather in the proper understanding of his object. Sedlmayr renders a judgment on the façade of San Marcello, writes at length on the façade of San Biagio in Campitelli, and comments on the plan for a church in the Colosseum.

Façade of San Marcello al Corso, Rome, 1681—1687, designed by Carlo Fontana. Engraving by Giovanni Francesco Venturini from Giovanni Giacomo de Rossi, Insignium Romae Templorum Prospectus (Rome, 1684).

With his remark that San Marcello represents “the most agreeable rendition of Bernini’s widely circulated concepts,” Sedlmayr does not go beyond the superficial judgments that have been made of Bernini’s closer and more distant successors over the past hundred years. Fontana is in many cases—as Coudenhove-Erthal also fails to note—an interpreter of very particular ideas and projects of Bernini’s, but it is precisely from this point of view that the structure and façade of San Marcello cannot be understood. Let us read what Coudenhove-Erthal has to say about this façade, on pages 52-ff. He attempts to do justice to the specific character of the work, without overlooking the fact that “the individual formal elements are in no way original.” It is unfortunate that Sedlmayr’s remarks flatten such distinctions. Sedlmayr might instead have deepened and indeed corrected Coudenhove-Erthal’s analysis. It would have been possible to show (in contrast to Sedlmayr’s judgment) how Fontana’s architectural thinking in the façade of San Marcello is oriented to the façades of the late 16th and early 17th centuries (for example Santa Susanna), in its working-through of the problematic style of Bernini and the high baroque. This interpenetration of different principles of design unified in a synthesis with various novel accents, which characterizes the Roman late baroque, achieves its first valid expression as a church façade in Fontana’s design for San Marcello. The façade’s enormous influence rests on this exceptional position. This influence would have been impossible if here it were only a matter—as Sedlmayr claims—of a pleasing combination of retrograde elements, rather than of a form with its own original stamp. Of the many façades connected with San Marcello, let us only recall Juvara’s façade of Santa Cristina in Turin.

The recently dismantled façade of San Biagio in Campitelli is at the center of Sedlmayr’s considerations. Here he finds ideas “that are among the most modern and essential between Borromini and Fischer von Erlach.” Because Sedlmayr gives a detailed explanation of his point of view—which is convincing at first glance—we must examine his argumentation. What is new and consequential about this façade, according to Sedlmayr, is that it offers both a near as well as a distant view: “What is decisive for an understanding of this work is to see that the upper story is another spatial unfolding of the lower story and that, for a more distant beholder, this upper story merges into a repetition of the lower.” Sedlmayr’s genetic classification of this façade falls short, however, because this correspondence of the lower to the upper story does not in fact occur in the distant view. Sedlmayr meets the objection that a distant view is entirely impossible—and, as maps of the city reveal, could never have been possible—with theoretical considerations that do not stand up to serious inspection, and which we can refute with Sedlmayr’s own words in his argument against Riegl’s interpretation of the wall of the Laurenziana: “in that place a view at a distance is not even possible.”3 Beyond this observation, it can be theoretically demonstrated that the upper story cannot merge into a repetition of the lower story, given that in orthogonal projection the articulation of the lower story is only continued in the bent pilasters of the central avant-corps. Furthermore, the openings in the side elements of the upper story do not lie on the axis of the lower story! (These facts are also evident from the drawing in Rossi’s Studio d’architettura civile, which Sedlmayr cites to entirely opposite ends.)

View of San Biagio in Campitelli (center). Detail from Giovanni Battista Falda, Nuova piante et alzata della città di Roma con tutte le strade piazze et edifici de tempi (Rome, 1676).

To do justice to the specific structure of this façade, one must start with the “surrounding space that functionally belongs to it,” that is, from its position in a narrow alley. This location induced Fontana to conceptualize the façade with the view from the Piazza Aracoeli in mind. From here, the corner of the lower story joins with the recessed, concave section of the upper story (upon which hangs the Chigi coat of arms) in a closed and very finely calibrated plastic form. From this point of view, insofar as one simultaneously perceives the articulation of the central avant-corps which passes through both stories, it appears as if the purely frontal view will present the main aspect. Later, one learns that the frontal prospect cannot in fact be perceived at all. The artist thus neglected to unify the articulation of the façade for the frontal view. If it were possible to see the façade along its axis from a distance, the upper story would appear to be off kilter. This devaluation of the frontal view and valorization of the corner is likewise quite marked in details such as the volutes, which rise from the center to the corners. A detail such as this is incomprehensible if we accept Sedlmayr’s approach.

We must try to be clear about the significance of Fontana’s solution. All baroque church façades to that date had been axially oriented, even when their central axis could not be seen due to given spatial conditions. Thus, they were constructed according to an inner geometric-objective lawfulness, indeed to an extent without concern for the viewing subject. The above analysis of the relations between the body of the building, its surrounding space, and the viewer has shown that the façade of San Biagio, on the contrary, is conceived with only the viewer in mind.

This “psychological” attitude towards architecture (cf. Die Zeichnungen des Gianlorenzo Bernini, 102) is Bernini’s proper domain. Fontana had been exposed to this mode of thinking at the beginning of the 1660s through his work on Bernini’s projects for St. Peter’s Square. San Biagio is probably the first application of Bernini’s new attitude, which was to be so consequential for the later history of architecture, to the construction of a façade. I have thus arrived at a result that is quite the opposite of Sedlmayr’s: the façade of San Biagio is essentially a specific product of Bernini’s spirit, whereas that of San Marcello exhibits a particular and truly “Fontanesque” structure. (Compare, also, the dates!) Sedlmayr’s mistake is to attempt to deduce the principle of “shifting sight,” which he has introduced elsewhere, from the fact of the differentiated spatial development of San Biagio’s two stories. Here, for Sedlmayr, theory has become an end in itself; it is not derived from observation.

Sedlmayr makes a positive contribution to our knowledge of Fontana’s project for a church in the Colosseum. I will content myself with noting that Sedlmayr underestimates the importance of the “content of the invention” (a Christian church in the midst of ancient ruins) for an understanding of the project, whereas Coudenhove-Erthal bestows a proper appreciation on this aspect. Only the combination of these two analyses makes the artistic idea fully comprehensible.

Sedlmayr’s remarks tend to merely sketch out an image of Fontana’s art rather than to clarify it: via the projection of an alien structure, a typical work of Fontana’s youth (San Biagio) is accorded a genetic significance that it does not merit, while an independent work of the middle period that is of great importance for the historical development of architecture (San Marcello) is devalued.

In this dispute I am not concerned only to correct certain particulars: I also mean to point out the danger that arises in various methods of scholarship when dogmatists claim absolute validity for their findings.

It is not surprising to find that Sedlmayr concludes with curious suggestions for a “doctrine of value.” Although here too Sedlmayr, as always, brings forth ideas that are worth taking seriously, let the reader try to imagine the following actually carried out: “The idea of the whole would have demanded different moldings and detail motifs that one could practically derive from a careful apprehension of the larger motif.” If that were possible, it would extinguish the personal handwriting of the artist, his style. But this self-revelation of the Platonic idea of a work is in itself unthinkable. It is rather out of such stuff that the spirit of the reconstructive art historian invents his own work. However much an already-existent practical and emotionally-based approach to value may be in need of theoretical foundation, and however true it is that a work’s level of artistic quality need not coincide with its historical importance, this path nonetheless leads rather to an annulment of the individual than to a useable doctrine of value. Here, too, Sedlmayr ends up with what is clearly a reversal of the aim to which he aspires: theoretical exaggeration blocks his access to the individual work.

Berlin, October 1931

  • 1 I would nonetheless briefly like to address two important points. A basic and omnipresent failure of this monograph on Fontana lies in the author’s premature choice of his theme. A monograph ought not to have been devoted to Carlo Fontana until the Roman early and high baroque had been properly examined, that is, before studies had appeared of Maderna, Longhi, Cortona, and so forth; the Bernini material that I have published in collaboration with Brauer was also not yet available to the author. Inadequate economy of labor is a persistent failing in art historical research. In this case, the author has deprived himself of countless discoveries that otherwise might have been his.—Even if here we can only speak of shared guilt, his work nonetheless suffers from a further omission that ought not to have occurred. A Fontana monograph should base itself on the 27 volumes of drawings and manuscripts that King George III bought from the Albani collection for his library at Windsor. The author’s reason for this omission is all the more inexcusable given that he himself notes (on page 72) that a volume containing Fontana’s drawings for the Curia Innocenziana was lent from the King’s collection for a 1911 exhibition in Rome. It would have been easy enough to follow up on this catalog entry.
  • 2 [Translator’s note: Wittkower here refers to Sedlmayr’s “Zu einer strengen Kunstwissenschaft,” Kunstwissenschaftliche Forschungen 2 (1931), 7–32; available in English as “Towards a Rigorous Study of Art,” trans. Christopher Wood, in Wood, ed., The Vienna School Reader: Politics and Art Historical Method in the 1930s (New York: Zone Books, 2000), 133–79.]
  • 3 Sedlmayr, Die Architektur Borrominis (Berlin: Frankfurter Verlags-Anstalt, 1930), 152; quoted from: Hans Sedlmayr, “The Architecture of Borromini,” trans. Karl Johns, Journal of Art Historiography 14 (June 2016), 91. [Translator’s note: Sedlmayr’s critique is of Alois Riegl’s description of Michelangelo’s design for the façade of the Basilica di San Lorenzo, Florence, in his 1908 book Die Enstehung der Barockkunst in Rom, available in English as: The Origins of Baroque Art in Rome, trans. Andrew Hopkins and Arnold Witte (Los Angeles: Getty Research Institute, 2010).]
Façade of Santi Vincenzo e Anastasio, Rome, 1646-50, designed by Martino Longhi the Younger. Photograph from Rudolf Wittkower, Art and Architecture in Italy 1600 to 1750 (5th edition, New Haven and London: Yale University Press, 1982).

3. Hans Sedlmayr, “On the Concept of ‘Structural Analysis’ (Coudenhove-Erthal’s Fontana Monograph, Once Again)”

Originally published as “Zum Begriff der ‘Strukturanalyse’ (Noch einmal Coudenhove-Erthal’s Fontana-Monographie),” Kritische Berichte 3/4 (1930/31 and 1931/32), 146–160.

The comments on Coudenhove-Erthal’s Fontana book in the penultimate issue of this journal in a certain sense do not in fact constitute a “critical report.” They were originally intended to appear in a place where the justification of critical findings could be reserved for other publications. They were accepted by Kritische Berichte because the journal’s editors viewed the contribution as a possible and legitimate type of criticism, even in this form. That the findings with which I operate derive from completed structural analyses, in accordance with the demands that we ourselves have imposed, goes without saying. Wittkower’s objections grant me an opportunity to expand my text into the form I would have given it had it originally been drafted as a “critical report” and at the same time present an occasion to clarify the frequently misunderstood distinction between “good” and “bad” description (arbitrary description, structural analysis, style analysis) by way of a concrete example.

In my notes in the third issue of this journal, I attempted to transform the dominant understanding of Fontana’s art—which persists in Wittkower—through the thesis that Fontana “had ideas that are among the most modern and essential between Borromini and Fischer von Erlach.” Fontana’s creative powers should not be judged by San Marcello, but rather by two other works. This claim was obtained from structural analyses of these three works (among others). Wittkower disputes the correctness of the findings in two of these analyses. A response would be superfluous if it were not possible to reach a determination based on reasons. In what follows, I intend to show that this determination can be made unambiguously.

a) San Marcello. Wittkower disputes the correctness of the claim that the façade of San Marcello is a “rendition of Bernini’s widely circulated concepts.” Furthermore, he considers the statement empty; he believes that it is not based on concrete observation. In reality, the claim formulates quite specific observations; it is the result of a complete and detailed comparative structural analysis which my notes concisely summarize.1 To wit:

The façade of San Marcello is a late baroque transformation of the high baroque façade of Santi Vincenzo e Anastasio by Martino Longhi the Younger (dated 1650).2 It is indicative that neither Coudenhove-Erthal nor Wittkower have noticed this. These are poor marks for a critic of the “second” science of art. For, as the following will make clear, such a critic will have understood next to nothing about San Marcello if he is incapable of noticing this relationship.

In order to see the genetic relationship, one must first of all observe the three narrow and equally spaced orders (columns) that in Santi Vincenzo e Anastasio flank the middle section of both stories. This very unusual motif first appears—so far as I know—in Longhi. It returns in Fontana. Here, in the upper story, are a column and two pilasters in a plane, in the lower story three columns, but spatially staggered in a way that differs from Longhi, not thus:

but rather thus:

Later we shall understand the meaning of this divergence.

In the façade of San Marcello, Fontana transformed the Longhi façade in a double process: 1. The Longhi façade is broken into three clearly separated layers, each staggered behind the other and each the width of an order of columns (a module). These layers can be pulled apart neatly and without violence—which is not the case with Longhi. It would be possible to cut out each these layers from the ground plan (Coudenhove-Erthal, fig. 17) in closed, curving slices of the scissors.

2. The resulting façade as a whole is made concave.

Let us in imagination take apart the layers of this façade. The foremost layer consists of two full columns on either side of the axis together with the window frame over their center. Remove this layer (which could very well exist on its own as an independent structure): what now appears as the next layer is a two-story façade, framed on either side by a column and two pilasters; here the upper and lower stories are identical. The third layer is encompassed, on the lower level, by the single-span bays, each of which is bordered by a single pilaster on the side, and on the upper level by the palm volutes and their connected pilasters, above which lies an entablature.

Let us presume that Fontana intended this “modulation” in clear layers. Let us furthermore trace, step by step, this façade’s transformation of Santi Vincenzo e Anastasio. This will be a little tedious, but in the process we shall learn something that can be learned no other way. (1) Longhi’s triple-column group must be split into two groups to prevent the foremost layer from becoming overly weighted: thus, either two columns in the foremost layer and one in the second, or the reverse. What was executed was the first option; one need only visually imagine the second to understand why it represents an inferior solution. (2) If there are two columns in the first layer, in the second layer behind the columns there must accordingly be pilasters on the lower level—and thus likewise above. Above the third column, which already stands in the second layer, there must also be a column in the upper story. One can now understand why it is that in the upper story two pilasters must stand alongside a single column. (3) Behind the columns of the second layer a pilaster must stand in the third layer. (4) On the side, the boundary of the third layer must be a pilaster. If there were a column here, a fourth layer would have to be introduced, which in the elevation would then no longer stand out from the third layer.

I know that such analyses have been decried as “rationalistic,” but in this case it is Fontana who is all too rational. There are countless structures for which an attempt to provide such a rational motivation for the individual motifs would remain fruitless. One of Fontana’s distinguishing characteristics is to have fully “apprehended” this logic of the form-making process, which is distant from all intuition [diese anschauungsferne Logik des Gestaltungsprozesses].

In order to see more clearly the modification that has here taken place, let us consider once again Fontana’s point of departure: the façade of Santi Vincenzo e Anastasio. Here the columns in both stories stand in a single plane, which twice bulges out only slightly around the center—for precisely which reason this slight staggering has an enormously insistent aesthetic effect. The wall behind the columns is not in the same sense a second relief layer, as are the layers of San Marcello; it is rather a foil for the colonnades, in a similar (though by no means identical) sense as is the wall behind Hellenistic colonnades. They consequently also do not incorporate any plastic motifs (niches with figures) in the side fields; they are incapable of doing so. The emptiness of these side bays, which exerts a powerful aesthetic effect, constitutes a “silence” that contrasts with the drone of the three-column motif in the center. In general terms, the façade is a flattened Hellenistic portico (“flattened” here meaning: with little spatial depth to support the columns), with the columns arranged in front of the wall in a manner that is plainly bizarre in terms of antique “rules.” For a sensitive viewer, I believe, a curiously fractured echo of the antique-Hellenistic timbre is clearly to be perceived even in the first unarticulated overall impression.

This structure emerges in a process that is entirely different from that of San Marcello. Longhi starts with a received façade schema (it is not necessary here to think of any specific structure), but from the beginning the development of this schema is determined by the powerfully intuitive vision of the magnificent column-triad and the three nested gables; over its entire course the design does not for a moment deviate from this primary vision. Fontana, by contrast, starts with an entirely concrete structure and transforms it according to an abstract ideal (separation into clear layers). His revision of Longhi’s façade is more ingenious, more “spirited,” more interesting, but also infinitely poorer in visual content. It feeds on the substance of the older structure. One could put it this way: Longhi is more creative than Fontana. But Fontana indeed comes off particularly poorly if we start with San Marcello.

How then can one describe the façade of San Marcello as a “rendition of Bernini’s widely circulated concepts”?

The tidy staggering of layers, the thickness of which is determined by the module of the orders of the columns, is one of Bernini’s formal principles (see his tower designs for St. Peter’s, the altar in Santa Maria della Vittoria, and the gate of the planned stables on the Quirinal,3 among others). Compare Borromini, Cortona, or the early baroque (Maderna: Santa Susanna). It is linked to the dominanttendency in Bernini’s design, which we usually—far too generically—label as “classicizing.”—Making this form concave is one of Bernini’s typical operations, the significance of which it is unnecessary to indicate here. Fontana produces the final façade of San Marcello through a regrouping of the Longhi façade, precisely as Bernini does when he layers older forms in a series of planes, for example in the abovementioned portal for the stables on the Quirinal. The foremost of the three layers is practically a Bernini altar, as H. Bergner has already noted, incidentally (Das barocke Rom, 1914). Finally, the substrate upon which Fontana works these Berninesque operations, Longhi’s façade itself, is likewise a highly independent product of the circle of ideas around Bernini. The preference for freestanding columns is characteristic of this group. It is connected with the adoption of Hellenistic ideas, which is otherwise foreign to the Roman high baroque (the latter is in general oriented to the relief of the ancient Roman wall orders). All of which could be shown in detail. This is the basis of Bernini’s relative proximity to contemporaneous French architecture. The colonnades of St. Peter’s are unthinkable without an orientation to Hellenistic architecture.4

The above claim is thus well-grounded. Neither is the epithet “most agreeable” arbitrary. Fontana’s façade brings these ideas, which by 1680 had already become rather commonplace, to a polished academic form. Their apparent richness is fundamentally schematic and easily accessible. A high baroque idea is on the one hand (through the concave curvature) rendered more baroquely complex in an easily understood fashion, while on the other hand (through the articulation in clear, comprehensible layers) its structure is simplified, explicated, and strictly logicized through its realization. The façade—which is uncommonly revealing for this phase of Italian architecture—offers the prime example of an academic version of the baroque that has been distilled to simple principles.

b) Good and bad description. And let us now once again read the above description of San Marcello and, following that, Coudenhove-Erthal’s. (His description, like ours, is comparative: first he describes what San Marcello has in common with the façade of Santi Margherita ed Emidio,5 and then what distinguishes it.)

Is there really no difference between these descriptions? Or is it not rather the case that in the first description the structure of the form is clarified even down to quite concrete details, whereas in the second case the function and connection of the motifs remain unclarified? The answer need not result from any partiality: it follows from the matter itself. For indeed the first description accomplishes something that the second does not, and to which it cannot oppose anything equivalent.

From the first description, for example, one immediately understands the function of the curious frame over the portal, which has unsettled viewers from the beginning.6 It is in a certain way a “grip” for the eye, a handle, that allows the foremost relief layer to be seen more easily in its separation from the other layers—thus, it refers to the fundamental principle. For this reason, it intentionally overlaps a form in the second layer lying behind it. In Coudenhove-Erthal’s description the concrete function of this motif is incomprehensible. I have furthermore shown how the location of every column and every pillar can be unambiguously determined on the basis of these basic concepts, which we now understand very well. A structure developed, as Coudenhove-Erthal suggests, from the earlier façade of Santi Margherita ed Emidio would resemble San Marcello only in its most general schema and would remain entirely undetermined in all of its concrete particularities.

This, however, is what is essential to a structural analysis: to let the concrete form of the artwork in all its details emerge in an evident progression from a grasp of the evident basic conception (and from the hierarchy of further conceptions of second and further orders).7 Hence the above description is only the beginning of a structural analysis which can be developed further and correct itself through the discovery of additional formative principles [Gestaltungsprinzipien]. Coudenhove-Erthal’s description is however not a structural analysis, first by reason of its result, because it does not accomplish this reconstruction [Nachgestalten], and secondly by intention, because it knows nothing of this demand.8

Hence it is unfair to Coudenhove-Erthal’s description (which, in a way quite typical for the “first” science of art, loosely unites individual observations and indeed possesses real meaning and value within these limits) to pretend that it is a structural analysis. For if we were to approach it as structural analysis it would be a prime example of a “bad” and abstract descripttion.

It is abstract in a double sense. The object with which the façade is compared is far too distant, such that the commonalities become both inessential and non-evident [unanschaulich]. And the markers of distinction (the differentia specifica) are statements of far too great generality. This is not only a logical failing of the method, but also a symptom of the fact that the specificity of the façade has been insufficiently grasped in vision [im Sehen], that it is founded on a confused and indeterminate perception [Anschauung]. It would be possible to foist a form upon the description that would have only externalities in common with the described form, and this characterization would still apply.—The description is “bad,” however, not because it consists of false statements—there are indeed correct statements among them9—but rather because the individual findings, both true and false, essential and inessential, very general and very specific, stand unconnected and unarticulated beside each other. Here, too, the error lies not in the description as such, but clearly in the perception upon which it is based: it has not achieved “formed seeing” [gestaltetes Sehen]; the perceptions remain fragmentary.

This ought not to be taken as an indictment of Coudenhove-Erthal: I do not believe that his intention was anything other than to communicate a loosely ordered series of individual observations that seemed important to him. It is the task of the second science of art to sieve out whatever is useful here for the understanding of the form. Coudenhove-Erthal makes no claim to deliver the second science of art.

With his claim that Longhi, Maderna, etc. ought to have been researched before turning to Fontana, Wittkower inadvertently demonstrates that this distinction, which he does not wish to admit, is real,10 that every problem looks differently when seen in the light of either the “first” or “second” science of art. His demand for an “economy of labor” makes a great deal of sense for the “first” science of art, if we guard it from a recursus ad infinitum by way of a better formulation; it is quite meaningless for the “second” science of art. For the latter there can be only one demand for economy of labor: to begin where one has found an access to the concrete understanding of the form. And without this understanding, all 27 volumes in Windsor will be of no help in comprehending Fontana’s art.

The second science of art establishes rational criteria for this understanding, with the aid of which it judges whether in a particular case we have attained understanding (or at least an approximation of understanding) or not. The first science of art, and Wittkower along with it, knows only the criterion of immediate self-evidence. Thus, it raises precisely that claim to “absolute validity of their results” that Wittkower (thanks to an ideology that does not match his actual practice) rejects in theory.

Where a concrete evident understanding [konkreten anschaulichen Verstehen] has not been achieved, statements about the similarities between forms do not touch upon the essential, but only upon externalities. Equivalencies are seen between forms that structurally have nothing in common; on the other hand, similarities between superficially dissimilar forms go unrecognized; as a consequence, “history” too is written in these terms. Hence neither Wittkower nor Coudenhove-Erthal have noted the relation between San Marcello and San Vincenzo; this also explains the fact that Wittkower tries to derive the façade from the façades of the early 17th and late 16th centuries. The similarity here is only external, namely the staggered articulation of the walls towards the center; the manner of the staggering and its meaning is entirely different. This can be seen especially well in Santa Susanna, which Wittkower also cites, and can also be seen perhaps even without any special prompting in San Marcello.

c) San Biagio in Campitelli. Even if my interpretation—to which I am committed—proves to be false, it is nonetheless an attempt to understand a motif that Wittkower and Coudenhove-Erthal have completely overlooked: the oblique jambs and the sinking entablatures above the doors and windows.11 From a near or plastic position [bei einer nahsichtig-plastischen Einstellung] this motif is meaningless, as it enforces an optical-distant [optisch-fernsichtig] approach upon the viewer, both in the corner view as well as in the frontal view. This is an absolutely uncontestable fact, for the details are already present in Rossi’s print. In the upper side field of Coudenhove-Erthal’s plate 2 one can see that the perspective effect functions correctly in the distant view.

An analysis that neglects this unmistakable motif is baseless.

This observation alone immediately refutes Wittkower in two points: 1. The main view lies on the axis of the building, and 2. This view demands a standpoint far in front of the façade, which cannot be occupied in reality.

1. The perspectival system of the façade is not unified. The three lower niches and the upper central window are based on a standpoint aligned to the frontal axis of the façade, while the two upper side windows are based on standpoints along lines that intersect the main axis at 45 degrees. This shows that the corner view was intended (Wittkower does not mention that I also consider this view possible, but only as a mere “initial view”), but at the same time proves that a distant frontal view is demanded and that the latter represents the main view. This is already evident from the fact that, in perspectival terms, of the six bays of the façade, four are frontally oriented and that these could exist on their own as an independent façade, whereas the two diagonally oriented fields could not.

Wittkower’s claim that “The artist thus neglected to unify the articulation of the façade for the frontal view” is simply false, for the repetition of the oblique corners in the left half of the façade—which cannot be seen at all from the corner view—shows that for Fontana everything depended on producing frontal symmetry. On the basis of Wittkower’s description, one might conclude that the building displays an “oblique corner” only towards the Piazza Aracoeli. San Biagio thus falls precisely into that group of buildings from which Wittkower aims to exclude it: “All baroque church façades had been axially oriented, even when their central axis could not be seen due to given spatial conditions.” This is just the case with San Biagio. (Incidentally, the antithesis, as Wittkower formulates it, is inadmissible.)

The claim that “If it were possible to see the façade along its axis from a distance, the upper story would appear to be off kilter” is misleading: the upper story as a whole would not appear to be off kilter, but rather only the two bays above the two corresponding bays in the lower story—and even these only under certain conditions (see below).

2. A distant view of the main front is however not only impossible in the narrow alley but, as Wittkower correctly notes, could never have been possible. What should we conclude from this? Various hypotheses are imaginable, of which only two can be seriously considered: a) Fontana expected that the area around the façade could subsequently be laid open. We can surely exclude this assumption. b) The façade was composed for the distant view and was only subsequently transplanted into this context. This is conceivable for Fontana’s mentality, but it implies that it is necessary to remove oneself in imagination to a distant standpoint, just as I have described. It is true that this view from a standpoint that is impossible to occupy in reality is highly abstract, but here too the abstraction is Fontana’s and not the viewer’s, whose conduct is indeed practically dictated here.12

One task of a complete structural analysis, which due to lack of space can here only be reproduced in extreme abbreviation, would be to describe what we see from the various standpoints on offer (fig. 4).

1. Corner view from the Piazza Aracoeli. That it is in no way possible to integrate visually what to Wittkower appears to be “a closed and very finely calibrated plastic form” is self-evident from the schematic drawing in figure 1. This closure simply is not present; of the three fields, two have a different perspectival orientation, and the third is neutral. No architect of the 17th century would have seen anything here but a mere “non-form.” The outlook that Wittkower recommends has perhaps only become available to painters since the end of the 19th century.

The building itself prescribes the correct articulation: We see a field in the upper story that faces us directly and four fields that turn to the alley (as is unmistakable from the sinking entablatures of the windows) and which together constitute a façade of the well-known type “three fields below, one in the center above.” Finally, there is an inert upright field in the upper level (see figure 2 and Coudenhove’s plate 2).

The concrete function of the volute motif can only be understood as follows: It disaggregates the unity of the diagonal corner field and transfers this unity to the other two fronts: the main front and the perpendicular side front.

2. View from a fictive distant standpoint in front of the façade. Rossi’s elevation can serve as a substitute for this view. It is correct that objectively speaking the windows of the upper side fields do not sit on the axis of the corresponding lower windows. But the way in which the form appears from a distance cannot be deduced from this objective state of affairs. Experiments with trained observers show that this slight and, moreover, concealed deviation goes unnoticed unless attention is somehow drawn to it. But this alters one’s approach. Viewers who initially saw the façade as represented in figure 3 now instead see it articulated as in figure 4. This is just the same as in the case I have offered as a paradigm for “shifting vision” [verwandelndes Sehen]—the high altar of the Franziskanerkirche in Salzburg.13 There, too, from the initial standpoint at the entrance of the church one does not necessarily see only an oval relief; rather, two main possible viewpoints are present beside each other and are, in a sense, confused. From a closer standpoint one view falls away, by contrast. In essence it is just the same here.

We are still far from concluding the analysis of this structure, however. There is a third possible viewpoint. We see it alongside the other two when we try to imagine how the form of the façade emerged in concreto. There are (at least) three options: 1. From a façade of the type that, for example, figure 4 represents, minus the shaded areas, by an addition of the two oblique corner fields. 2. From a façade of the type of San Gregorio in Monte Celio (see Rossi), by “folding back” the two upper corner fields. 3. From a building of the type of Santa Maria di Loreto, by splitting off one “third.”

Hans Sedlmayr, schematic diagram of San Biagio in Campitelli.

The structure of San Biagio is so complex that it is certainly possible to contest my findings. But one can contest them only with arguments derived from structural analyses in which at least as much of the actual reality of this work is drawn into the interpretation as in the attempt at an interpretation that one is contesting.14 What Wittkower provides (and what he attributes to Coudenhove-Erthal) is in no sense a structural analysis. It is indicative that Wittkower only once (and Coudenhove-Erthal at no point whatsoever) attempts to justify the adequacy of his view on the basis of the structure actually in question, namely, when he (to an extent correctly) interprets the function of the corner volutes in San Biagio. But here, too, one particularity is simply brought into connection with another, rather than all of them being brought into connection with their common center.

d) On the design for a church in the Colosseum. Wittkower has here misunderstood a term. The “content of the invention” of which I spoke naturally has nothing to do with the “objective meaning” of the structure15 or the “allusion” that it contains (a Christian religious building in an ancient circus!), but rather lies in the idea of placing a central plan building eccentrically in the crater-like space of the Colosseum.

e) Wittkower ranks the works here under consideration according to their significance for the development of Fontana’s personal style. San Marcello seems to him more important because it is a more perfect expression of the real Fontana.16 But this is a far too limited standpoint from which to view Fontana’s historical significance correctly. It does not illuminate the relation between Fontana’s ideas and the most essential new tendencies in Europe at the time. (I have already spoken of this, however, and what I have said cannot be refuted by a retreat to a more restricted line of questioning.) It is indeed the case that San Marcello is more characteristic of the version of Fontana that represents the Italian late baroque. But precisely for this reason it represents all the defects of this epoch, which was to see Italy reduced to a secondary role in European architecture. And if we concede Wittkower’s claim that San Biagio is “a specific product of Bernini’s spirit,” this nonetheless would apply not to the structure (in which Bernini’s ideas undergo an inflection that goes far beyond anything found in Bernini himself); what is correct is that in San Biagio there remains something of the Europe-wide validity of Bernini’s ideas, something that the Roman late baroque would almost entirely abandon. A more concrete history of art cannot hide these distinctions behind an outlook oriented purely to style.

The special position of San Biagio and the illustrious future of the ideas that are here only lightly touched upon (the effect of which was not, as with San Marcello, to end with a few sterile modifications) is—even if we disregard the results of the structural analysis—already evident from an unarticulated overall impression of the building. San Biagio is perhaps the only building of the period in Rome that—with minimal alterations—could have been created in Austria around 1700. Fontana himself was not capable of grasping the significance of his concept and the potential for its further development. It is only in the late baroque system that such ideas were to be guaranteed an organic incorporation and creative development. It is not by chance that the particular motif of the concave corners, which practically demand projection outwards from the line of the wall (this is absent in San Biagio), only enjoyed its productive unfolding in the freestanding palaces of the German late baroque.

f) On a future doctrine of value. Without a more concrete understanding of the works, a doctrine of the value of artistic forms cannot be developed or even productively discussed. The sentence upon which Wittkower seizes, however, relates to such an elementary and simple state of affairs that recourse to a developed doctrine of value is wholly superfluous.

I have, in essence, claimed that in San Biagio Fontana was unable to discover a “fitting” design for the detail motifs. Wittkower does not dispute the correctness of this statement, but rather argues that the statement is meaningless: that what the statement refers to cannot exist at all.17

The “unthinkable” claim comprises an observation and its interpretation. The observation consists quite simply in the fact that certain details are out of keeping with the whole which they organize. This is however an assessment of the sort that Wittkower’s art history likewise knows very well.

The observation can be interpreted in various ways: the discrepancy can, for example, be accounted for by the later modification of the details, or even a deviation from the architect’s design during construction (meaning that the architect did not prescribe these details), or indeed by any number of further causes; that in such cases a loss of quality results from this “lack of fit” is likewise generally accepted in the older art historical literature. But another interpretation is just as possible: the architect did not succeed in “finding” those details that would have brought his own idea of the work to an adequate form. The experience of anyone who is actively productive in a given field speaks in favor of the possibility of this interpretation, and it cannot be refuted with deductive, speculative arguments.

Wittkower will not permit this interpretation because it “would extinguish the personal handwriting of the artist, his style.” But this is of course by no means the case. For in order to arrive at this interpretation, it is indeed necessary to have previously determined that these moldings allow us to recognize Fontana’s “handwriting,” his style, to begin with, that is, that Fontana conceived them for this building.18

Just as, when the moldings are by someone other than the designer of the larger motifs, one may attempt on the basis of the work’s overall idea to determine which details would have brought this idea to its most concrete shape, one can likewise attempt this reconstruction when the same artist is responsible for both the details as well as the cardinal motifs. There is nothing “unthinkable” about this. It is indeed possible to understand even the imperfections of a work through an analysis of its structure.

g) Style analysis and structural analysis. Wittkower’s refusal to concede this fact is only comprehensible in terms of certain presuppositions of which he himself is unaware, and on the basis of which he judges and condemns. The most essential of these is evidently that the examination of artworks must consist in an investigation of their “style.” This is the fundamental thesis of the “abstract history of style.” From this point of view, the aim to go beyond a consideration of style through structural analysis must appear presumptuous.

An examination of an artwork’s “style” is not concerned with the totality of its qualities, nor with the relation and function of its so-called “elements,” nor with its concrete “shape” [Gestalt], but rather with those qualities that are “characteristic for” a group of forms [Gebilden].19 The groups in question are thus defined by their common origin, whether in an “era,” a “people” or “tribe,” a “landscape,” a “generation,” or an “individual,” etc.20

The artwork is here not seen as a “world” but rather as a medium through which something else “expresses” itself. It is thus not seen as an artwork: for the more one focuses on the artwork itself, the less one notices anything “characteristic” about it. In this view the autonomous artwork is degraded to something like “handwriting.”21

This approach must regard as “curious” the notion that an artistic idea (as soon as the artist, through whom and in whom alone it exists, decides to create a concrete work) itself poses demands that may be perceived or unperceived, realized or unrealized; its mode of viewing the work is entirely incapable of attaining to any knowledge of such things.

With this, the dynamic element of the work, which accounts for its value or lack thereof, likewise falls away, just as does that which constitutes the artist’s achievement: his “struggle” with the idea.

For the abstract history of style, value judgments arrive separately from statements about style. Style can be charted in bad works just as well as good ones. The interpenetration of value analysis and structural analysis (and thus the notion of a value structure) is foreign to it; in the case of Santi Vincenzo e Anastasio we have already seen how determinations of value are implied in the description of structure and cannot be extracted from the latter without vitiating the understanding of the formal structure.

This distinction in interest can easily be grasped in the different attitudes towards the individual artwork. With San Marcello, for example, a comprehensive analysis of style would aim to determine which characteristics express the specifically Italian character of the work, whether and in what traits it is “Roman,” in what ways it belongs to the period style of around 1680, whether it displays features that allow it to be recognized as a product of the artistic generation of about 1630, what it has in common with all other works by Fontana, in what respects it is characteristic of the middle period of Fontana’s career, etc.—for there are still more questions that can be posed in terms of “style.” In the end, the unified work would be fragmented in many directions, into many overlapping stylistic circles.—This is a completely different procedure from the structural analysis that I have sketched out above, which does not dissolve the unity of the work—but it is also something different from Coudenhove’s “loose” manner of description by the accumulation of individual observations, which suffices for Wittkower.22

Style analysis will always find it tempting to judge the quality of an artwork by whether or not it manifests a “pure” style. In order to state artistic distinctions of value, it must—temporarily—abandon its own way of looking at things and instead take up an entirely different mode of approaching a given form. In stylistic history, value judgments very often arise from the viewer’s private aesthetics.

It is clear that a history of art based on structural analyses will be quite different from one that begins with analyses of style. For indeed the abstract history of style, to which we have gradually become accustomed, emerges when—instead of first digging into individual works—we as quickly as possible seek to find a “general object” about which historical statements can be made, an object which is analogous to the “species” of “natural history.” Within a more comprehensive history of art, a focus on style is only one of the possible and necessary approaches, and indeed a peripheral one. It has a great practical significance because it enables attribution,23 even where no documents are on hand (this practical interest accounts for its early and rather over-hasty development), and it has considerable propaedeutic significance for a true science of art. But it has no significance in itself; if one practices it for its own sake one necessarily moves from art history to the history of spirit [Geistesgeschichte]24 (whether of the “Zeitgeist,” or the “spirit of the people,” the “spirit” of a landscape, of a generation, or of an individual person). The course of historical development has shown this, too; in terms of the history of scholarship, the phenomenon of the “history of spirit” is the complement to the abstract history of style. It is for this reason that the two are so compatible, and it is also for this reason that both find it so difficult to acknowledge the meaning, or even the existence, of a “science of art as a science of art.”

These brief characterizations should not be taken to devalue the “abstract history of style” as such, but rather only to counter its claim to be art history itself. The history of style has achieved great things for the initial development of a true science of art, which so to speak has come into being under its mantle and without any awareness of being something separate from it. But the science of art can only attain its complete unfolding when it disengages itself from the history of style, which at the present stage of its growth does more to hinder than to help it. This disengagement is now underway, under crisis-like conditions. The maieutic function of critique is to accelerate this process, insofar as it makes clear the embryonic existence and the potentialities for the development of a true history of art in distinction from the history of style. To this end, it must dispel the confusion that arises from the fact that the seeds of a genuine appreciation of art bear the same names as the terms of the “analysis of style.” The name of “structural analysis” (analysis of artistic structure) is meant to accomplish this separation; it can at any time be replaced with a better terminology.25

Wittkower’s anti-critique is welcome because it helps to refute many reservations held against the necessity of “theoretical” discussions. The case of San Marcello alone has shown that what one hopes to learn from the individual artwork is by no means a matter of indifference (and is not in all cases the same). But because this is the case, it is no longer possible to declare theoretical discussion unnecessary. When Wittkower calls these disquisitions “dogmatic,” he has things exactly upside-down. It would be far more appropriate to call “dogmatic” an attitude that proceeds from unacknowledged presuppositions. All of our so very unpopular efforts have as their goal the nullification of these “dogmas”—those of chronological history as well as those of the history of style—inasmuch as we put them to the critical test. “Theory as an end in itself” is unknown to us; theory is an unpleasant necessity. It is within the power of our opponents to make theory superfluous by moving to an approach that would more properly bear the name of a science of art. Then, one could restrict theoretical considerations to questions of technical research and method, whereas today—unfortunately!—it is still principles that are in question.

Vienna, December 1931

Detail of the façade of San Biagio in Campitelli.
  • 1 I must be permitted to oppose to Coudenhove-Erthal’s description, which Wittkower approves, my own description, and not only, as Wittkower would have it, its abbreviated result.
  • 2 A barely adequate photograph can be found in: Brinckmann, Handbuch der Kunstwissenschaft, plate 117; rough plan and elevation in: Domenico de’ Rossi, Studio, vol. 3, plates 39 and 40.
  • 3 Illustrated in: Wittkower and Brauer, 171b.
  • 4 Wittkower however denies this (Zeichnungen Berninis, 102): “The impetus for the conception of the freestanding colonnade came from quite a different direction than that of antiquity.” During the “Corpus domini” festival a long, covered corridor was erected in front of St. Peter’s by stretching a cloth between tall poles. “Bernini’s colonnades are fundamentally nothing other than the eternalization in stone of this makeshift canopy.” This claim, which is reminiscent of the fable of the origin of the Corinthian column, cannot replace a genetic derivation. Even if Bernini really was “stimulated” by this improvisation (which has not been demonstrated), he nevertheless made out of it a Hellenistic colonnade, and was quite conscious—as Wittkower himself notes—of ancient structures, which in any case are incomparably more similar to the executed colonnades than the canopy. (Furthermore, it is quite probable that with his square Bernini intended to outdo the most splendid square of ancient Rome, the Forum Trajani.)
  • 5 [Translator’s note: Now more commonly known as Santa Margherita in Trastevere or Santa Margherita di Antiochia. Fontana designed a new façade for the church, finished in 1680.]
  • 6 A relief was meant to be inserted in the frame. Diario del convento, published by Lina Muñoz Casparini, San Marcello.
  • 7 [Translator’s note: In the original, the italicized clause after the colon reads “aus der anschaulich erfaßten Grundkonzeption […] die konkrete Gestalt des Kunstwerks Schritt für Schritt bis in alle Einzelheiten hinab in einem anschaulichen Progreß entstehen zu lassen.” There are a number of difficult words here. In particular, anschaulich, “evident,” could also be translated as “visible,” “clear,” or “comprehensible.” It derives from the verb anschauen, “to look” or “to watch,” and is a key term in Sedlmayr’s aesthetics.]
  • 8 And thus, the so frequently misunderstood question of the “first” and “second” science of art is put to rest. Naturally every statement on the artistic aspect of artworks [das Künstlerische in Kunstwerken] (apart from negligible exceptions) ultimately aims at a “second” science of art; whether the second science of art is reached however, depends on—besides the viewer’s capacities—the demands that are made of these statements. We claim that it is not sufficient to “honestly attempt” a “characterization” (Wittkower), but rather that the essential demand must be to understand the form, function, and relation of the work’s parts—and with the concept of structural analysis I have just offered a brief definition of what I mean by “understand” here.
    This stipulation is frequently misunderstood as a claim that we ought to reconstruct the process of creation through which the artwork actually came to be. But this is a task of an entirely different sort. The process of creation can display an entirely different order and sequence from that of the hierarchy of motifs in the finished work, the meaningful interrelation and inner function of which is our sole concern here. (“Functional” elements of the form need not be identical with “genetic” elements.)
  • 9 It does not indeed suffice simply to demand correct statements, for it is quite easy to move exclusively within the realm of undoubtedly correct clauses. It must rather be demanded that the statements designate something essential about the object. And objective criteria for what may be seen as “essential” must be developed.
  • 10 Granted, it has no reality in the external sense. From the start I precluded the widespread misconception that, alongside the existing first science of art, a second science of art must be founded through my affirmation that “it is only permissible to speak of two sciences of art as a fiction: in reality they are two isolatable elements of a single ideal science of art.” (Kunstwissenschaftliche Forschungen 1, p. 8.)
  • 11 This is also how I designate the blind niches.
  • 12 Wittkower attempts to refute me with an argument that I myself have used to argue against Riegl’s interpretation of the distant view of the wall of the Laurenziana. But he overlooks that the wall of the Laurenziana precisely contains no such elements that impose a “perspectival” approach on the viewer.
  • 13 See: Sedlmayr, Österreichische Barockarchitektur (Vienna: Filser, 1930), 11–12.
  • 14 In principle, a non-distant view would also be permissible if it allows a different explanation for the emergence of the perspectival elements. This is what Wittkower would have had to attempt in order to counter my interpretation.
  • 15 Cf. “Die area capitolina des Michelangelo,” Jahrbuch der preußischen Kunstsammlungen 1931,    vol. 2.
  • 16 In Coudenhove-Erthal, San Marcello appears out of chronological order as the first plate. This can only mean that he sees it as Fontana’s masterwork.
  • 17 One can naturally object that the claim is nonetheless correct in this particular case. As counterevidence, however, I would need to offer analyses of the moldings and their function in the work as a whole, for which there is not enough space here.
  • 18 Quite apart from the fact that it is entirely characteristic of Fontana’s personality that he was not able to discover the “fitting” details, as is the particular manner in which he fell short.
  • 19 Whether the qualities to which one pays attention are integral stylistic characteristics or fragmentary stylistic markers is a question of a different order.
  • 20 In terms of form, one can say: The interest of the history of style is primarily classificatory; the point of view from which it classifies (origin in…) is strictly speaking extra-artistic.
  • 21 Cf. the already cited remarks in: W. Ziegenfuß, Phänomenologische Ästhetik, 102.
  • 22 In its essential points this critique of the abstract history of style follows from Benedetto Croce’s critique—which has received far too little attention in German art scholarship. By contrast, however, we have a different and less skeptical attitude towards the potential for a true art history.
  • 23 Taking this term in the most general sense, which also includes “allocation to an era” (dating).
  • 24 [Translator’s note: The word is more usually translated as “intellectual history” or “the history of ideas,” but these terms make nonsense of the following references to various forms of “spirit” (Geist).]
  • 25 If the first conscious attempts at structural analysis appear “formalistic,” one must keep in mind that at the start, in order to facilitate the new task, structures were intentionally chosen that display a “formalistic,” quasi-logical structure (San Carlino, San Marcello).—That a good description of structure must also consider the “objective meaning” of the work speaks for itself. (See “Die area capitolina des Michelangelo.”)