Structure/ornament and the modern figuration of architecture.

From: The Art Bulletin  |  Date: 12/1/1998  |  Author: Sankovitch, Anne-Marie

In architectural writings from about 1830 to the present a great many buildings are consistently described as divided between structures that represent one period style and ornament another. Why are Romantic historicists, early twentieth-century formalists, and contemporary contextualists all in agreement about the binary nature of these buildings? An examination of the literature on one such monument, St-Eustache in Paris, considers the covert, problematic function of "structure/ornament" as a spatially conceived narrative device; its relationship to "transitional" architecture; its (often unacknowledged) figurative significations; and its status in contemporary discourse as a discovery rather than a historically contingent invention.

Roman architecture often reduced the Greek orders to mere ornament applied to arcuated structures.

The Lombard chapel piled up ornament on the purist structure of the Florentine model.

In the nineteenth century a building was made a structure to receive an envelope of surface ornament.

To be authentically modern was to strip categorically from structure all ornament.

Few readers would find anything remarkable about the prominent use of structure and ornament in such statements, which resemble actual passages of innumerable modern writings on architecture. These two words seem to describe unproblematically only what is physically there; "structure/ornament" appears to embody the very nature of much built reality. We do not in general question, or even feel that it is necessary to question, what structure and ornament actually signify, or to ask why they so typically appear as an oppositional pair. Nor do we often seriously reflect on the historical origin of the pair (which is generally grossly misdated) or study the implications of that origination. In the absence of such critical analysis, we fail to realize how pervasive and compelling a figuration of architecture the structure/ornament pair is, and that it determines in massive ways much of how we think and write about many aspects of architecture and its history, and even to a large extent how we build. To initiate such an analysis is the primary aim of this essay, which is intended not to resolve issues attending specific historical sites but rather to excavate and closely scrutinize certain assumptions and problematics that pervade and frame structure/ornament, and thereby to put to critical questioning the seemingly transparent nature of much recent and current architectural discourse.

St-Eustache as Structure/Ornament Paradigm

Architectural history today frequently seeks to interpret buildings as objects shaped by and expressive of their social meanings and historical contexts. The function of a building is consequently understood as primarily representational and often as actively engaged in defining the social world of which it is a part. It would be both unexceptional and commendable to decide that the best way to grasp the realities of, for instance, a fifteenth-century Florentine church is to chart the competing economic, political, religious, and cultural forces that brought it into being and to interpret it as a material expression of the ascending wealth and status of the mercantile class during the period.

This alliance of contextualism and soft semiotics has been marshaled primarily as a reaction against the formalism that generally dominated architectural discourse from the late 1800s through the middle of the twentieth century and that coincided with modernism and its distrust of history. Since the embrace of social history around 1970, formalism and the internal history of architecture have been either rejected as elitist (or worse) or, more benignly, regarded as having discharged their necessary but narrow task so that we can now progress to a richer understanding of architecture in its full multidisciplinary complexity. In the efforts to anchor architectural form in its historical context, form itself has become self-evident and the procedures of formal analysis often tend to be taken as a given.

That a critical inquiry into the interpretive problematics of the properly architectural has been deemed irrelevant by many architectural historians is largely because the current revisionism has tended to restrict itself to questioning the scholarship of the earlier part of the twentieth century. Formalism is rebuffed because it is associated with an ahistorical approach, not because its procedures are inherently flawed insofar as strictly formal questions are concerned. The properly architectural is narrowly identified with the formal, and the latter is understood to be well understood.(1)

Modern strategies of formal analysis originated, however, not in the heyday of modernist formalism but far earlier in the historically attentive writings of nineteenth-century theorists. The Romantics and their contemporaries created a two-part model for interpreting architecture: buildings were located in the newly created, self-contained historicity of the evolution of architectural form, and simultaneously they were understood to be historically determined and contextually expressive objects. Architecture had its own immanent history, but this history was coordinated with social, economic, and cultural history. It was in the service of this dual project - not the prim, solitary demands of formalism - that new ways of conceiving and describing architectural form were devised. When, in the years around 1900, the historical part of this enterprise was suppressed, many buildings continued to be apprehended and described (if not comprehended) in fundamentally the same way as they had been for nearly a century. With the recent reemergence of history, many of the identical descriptions, with all their formal-historical baggage, are again being repeated, having tacitly if nonreflectively been granted apodictic status; prominent among these is the structure/ornament model.

The emergence in the nineteenth century of this immensely potent mode of architectural description and its ongoing reiteration through the present day can be illustrated in a brief survey of the descriptive history of one building, the church of St-Eustache in Paris [ILLUSTRATION FOR FIGURE 1 OMITTED]. In the first volume of the Dictionnaire raisonne de l'architecture francaise (1854), Eugene-Emmanuel Viollet-le-Duc wrote:

They wanted to apply the forms of ancient Roman architecture, which they knew badly, to the construction system of Gothic churches, which they scorned without understanding. As a result of this indecisive inspiration the large church of Saint-Eustache in Paris was begun and completed, a monument that is badly conceived, badly built, a confused heap of debris borrowed from all over, incoherent and without harmony, a sort of Gothic skeleton draped in Roman rags sewn together like the pieces of a harlequin's costume.(2)

Viollet-le-Duc's words are perhaps the most evocative rendering of a new visual and descriptive paradigm that configures St-Eustache as a building morphologically divided between its "skeleton," or structure, and its "rags," or ornament. Before the early nineteenth century such a two-part perception - even more, such a building - had been unimaginable.

When St-Eustache was constructed (1532-1640) and for some time thereafter observers were not much interested in allocating it a stylistic tag. Instead they saw (and esteemed) a monument notable for the abundance of its spatial and material traits: the great quantity and variety of its sculptural decoration, the great number of its piers and chapels, the great height of its vaults, and the unquantifiable spaciousness and richness of the building as a whole.(3) This was a superlative St-Eustache, which was seen, somatically experienced, and textually figured by the comparative grammatical framework whereby "big, bigger, biggest" or "some, more, most" equals "good, better, best."(4)

This "superlative" discourse was eventually displaced by the classical mode that emerged in France in the middle of the seventeenth century. The new discourse, which sought to separate the materiality of architecture from the idea it represents and to dissolve it into language, was highly theorized in its procedures as concerned both the creation and the apprehension of architecture. One interpretive gesture, however, was left free of theoretical elaboration, for it seemed self-evident: deciding to which of two possible manners, Gothic or classical, a building belonged.(5) This most apparently stable (because most reflexively deployed) gesture, this first casual glance, which effortlessly sees morphological traits that reveal the style of a building, proved imprecise and mercurial in the writings on St-Eustache. Everyone looking at the architecture of Paris "knew" that Notre-Dame and the Ste-Chapelle were Gothic, that St-Sulpice and the facade of St-Gervais were classical, but no such fundamental consensus was arrived at for St-Eustache. For some observers the building was Gothic,(6) for others it was classical or "modern,"(7) while for a third group it was both.(8) St-Eustache deflected the classical gaze and became an odd, unknowable building, inaccessible to the rational grasp of normal architectural discourse.

The confusion now caused by St-Eustache can be seen in Marc-Antoine Laugier's Observations sur l'architecture (1765). When he first writes about the building Laugier tells us:

The interior of this church is quite remarkable. The person who built it was strongly attached to Gothic architecture, and had a few feeble notions about Greek architecture. In this building he wanted to present some examples of the Greek orders. The result is those little columns hoisted up on excessively elongated pedestals, and which can be recognized by their bases, capitals, and fluting as belonging to antique architecture. This church marks an epoch in that it is only half Gothic, and, being like certain bordering provinces where opposing habits and languages intermingle, it signals the moment when Gothic architecture was about to die and Greek architecture was beginning to be reborn.(9)

In this partly Gothic St-Eustache, Laugier identifies classicizing columns that reveal themselves to his empirical scanning by their bases, capitals, and fluting. Although he describes only these isolated classical traits and does not indicate what about the building is precisely Gothic, at first reading his text seems to reveal a cognitively lucid, stylistically meaningful St-Eustache. But the building configured here is precarious, for its degree of Gothichess shifts as the text unfolds. First the church is strongly Gothic as the architect has merely a "feeble notion" of the "Greek" style, then it is half Gothic, and finally it is dying Gothic.

I would not insist on these distinctions, which follow a certain chronological logic and at least consistently describe the building as partially Gothic, were it not for a subsequent passage in Observations where a different St-Eustache appears, one that is entirely Gothic:

In our churches the vault is the principal object. It is there that the Gothic architect deploys his most brilliant resources. . . . In all the churches that we have built since the Renaissance of Greek architecture the vault is heavy and massive. . . . If one enters Saint-Eustache, there is nothing more elegant than the vault of this church. . . . If one enters Saint-Sulpice, there is nothing more insipid than that naked barrel vault.(10)

Now St-Eustache (diametrically opposed to the "insipid" classical St-Sulpice) is regarded as a characteristic specimen of Gothic architecture, a style that declares itself by the morphological feature of its distinctive vaulting. This abrupt visual realignment is accompanied by a historical repositioning of the building: from that moment when Greek architecture was first beginning to be born the church is pushed back to the time before this Renaissance. Is the St-Eustache of Observations Gothic and classical or purely Gothic? The text as a whole describes an elusive and changeable structure, a shifting, flickering architectural mirage where visuality is refracted and the most basic of epistemological assumptions called into doubt.

The nineteenth century brought St-Eustache to heel; architectural critics of the time, such as Viollet-le-Duc, now looked at and configured the building with a new architectural gaze, one that continued to search for style-revealing traits yet divided that recognition between the structure of a building on the one hand and its ornament on the other. It was just this possibility that the previous episteme was unable to entertain, and we should be careful not to endow a false immanent prescience on those classical texts that claimed St-Eustache was both Gothic and antique.(11) That the forms of a single monument could be composed of material traits from two distinct styles, with one category of traits coalescing into a building's physical structure and another into its ornament, was unthinkable and indeed was never stated.(12) Also, despite the fact that many eighteenth-century theorists (including Laugier) admired Gothic architecture or, more specifically, Gothic methods of construction, that admiration was limited to isolated motifs such as slender columnar supports, or to such resulting spatial effects as lightness and openness; it was not transposed to the recognition of a comprehensive tangible Gothic structure or skeleton in the sense that Viollet-le-Duc would imagine, either in Gothic monuments or, more to the point, in St-Eustache.(13)

It is only in the nineteenth century that a bipartite set of discursive spaces is produced, which all material architectural traits are seen to inhabit, variously and unambiguously, either as part of "structure" or of "ornament." In place of the oscillating, unstable St-Eustache that randomly proffered isolated details to the frustrated investigations of the classical gaze, a building of crystalline certainty emerges. Its morphology is no longer the object of uncertainty and, in fact, becomes a nonissue. The "structure/ornament" description seems to explicate and encompass the entire monument, apparently solving the mystery of the style of St-Eustache.

In the twentieth century this St-Eustache (either in its metaphorical guise of a clothed skeleton or its apparently literal one of an ornamented structure) is reiterated with the hallucinatory regularity of a mantra: "Saint-Eustache is a church with a skeleton of the Gothic type, overlaid with Renaissance adornment" (1910); "on a medieval structure there is Renaissance clothing" (1923); "the task of the church-builder . . . was to clothe a medieval skeleton in Renaissance flesh" (1926); "this new clothing covers an entirely Gothic framework of pointed arches and flying buttresses" (1944); "the medieval structure of this church is ornamented to the point of absurdity with elements in the Italian style" (1947); "this Gothic structure is . . . clothed in Renaissance forms" (1953); "to this medieval structure was unfortunately added decoration in the Italian mode" (1958); "evidence of the Renaissance style . . . is limited to ornament applied to the Gothic piers" (1978); "the whole church was submitted to the principle according to which Renaissance ornament was applied to the Gothic structure" (1984); "an Italianising ornamentation was applied to a Gothic structure" (1987); "only the decoration is representative of the Renaissance. . . . The structure is still entirely Gothic" (1989); "Saint-Eustache . . . is entirely Gothic in structure, although its decoration uses a classical vocabulary" (1997).(14) Furthermore, in the 1980s at least four authors cited Viollet-le-Duc's Dictionnaire description, allowing his words (which are irresistibly quotable) to corroborate or proxy for their own perception of the building.(15)

A St-Eustache is thereby produced that is virtually identical among the great majority of twentieth-century texts; a disarmingly simple building has taken root in contemporary scholarship with the tenacity of truth. How do we account for the strange success of this Gothic structure/Renaissance ornament St-Eustache, both as a construction in itself and as a phenomenon that continuously solicits duplication from the nineteenth century to the present day?

This question is not unique to St-Eustache, which is far from being the only premodern building that continues to be understood as divided between a structure that represents one style and its ornament another. Much of the architecture of sixteenth-century France and of Renaissance Europe outside Italy in general has been similarly configured by this binary concept. One author at the cutting edge of the antiformalist reappraisal of architectural interpretation describes a style of sixteenth-century Spain as "a hybrid local concoction of ornamental motifs applied without regard to the structure of the building," while another identified with formalist readings writes that in Germany "during most of the sixteenth century the Renaissance was simply a system of ornament . . . applied to Late Gothic structures."(16)

Nor does Italian Renaissance architecture necessarily escape the structure/ornament model. The Portinari Chapel at S. Eustorgio in Milan, for example, "represents . . . the transposition of the Sagrestia Vecchia of S. Lorenzo into the formal idiom of Milan. . . . The interior with its polychromatic blurring of the structure . . . and its prolific ornament . . . is far removed from the structural austerity of the Sagrestia Vecchia."(17) The Portinari Chapel differs from its transalpine colleagues mainly in that its blurred Florentine structure and its blurring Lombard ornament represent two regional variants of a single style, not two distinct period styles; yet like them it finds no place on the canonical Florence-RomeVenice axis and consequently is a building that formalism has perceived as marginal.

Moreover, since the early nineteenth century the history of architecture in general has become littered with more buildings that are, topographically speaking, marginal or peripheral and, temporally speaking, early, transitional, or late than buildings that have apparently achieved stasis at the central, high, or classical point of their style. In a great many of these cases the structure/ornament opposition is invoked as the buildings are described as formally cleft between two different period styles, different regional styles, or different phases of a single style.(18) Thus, onto the "Romanesque structure" of Bayeux Cathedral has been "grafted a heterogeneous collection of borrowings from early 13th-century Ile-de-France and English Gothic,"(19) and of the architecture of fifteenthcentury France it has been said, "Never . . . has Western architecture come closer to the luxuriant ornament of the East and to its fanciful profusion, which seems without purpose, and is certainly unrelated to the structure."(20)

Non-Italian Renaissance, non-Florentine quattrocento, French thirteenth-century architecture outside the Ile-de-France, and late Gothic architecture: it is precisely such fields, which were apparently misunderstood or entirely overlooked by formalists, where scholars have been particularly eager to follow the recent historical (re)turn in architectural interpretation. In the current climate a respectable argument can be made that the architecture of Renaissance Germany is not a lesser version of the Italian, not a marginal reflection of the center, but a historically legitimate phenomenon deserving of critical attention on its own terms. Similarly, a thirteenth-century cathedral that displays an early Gothic or Romanesque structure need no longer be disdained as provincially retardataire but can be interpreted as a declaration of regional identity, as responsive to the particular qualities of local building materials and masonry traditions, as having a contextually specific iconographic meaning or liturgical function, and so forth.

What present scholarship does not recognize, however, let alone encourage critical speculation about, is that such plunges into history often remain securely tethered to the peculiar revenantlike presence of the structure/ornament description. It is precisely this issue, and questions surrounding it, that I want to consider - that is, why nineteenth-century Romantics, early twentieth-century formalists, and contemporary contextualists are so frequently in agreement about the fundamental (structure/ornament) character of buildings about which they are otherwise in apparent disagreement.

The case of St-Eustache in modern architectural discourse is well suited to an inquiry into the problemafics of structure/ornament as a descriptive pair used to figure historical architecture. French nineteenth-century theorists were very much in the vanguard of the movement that created the new strategies for thinking about buildings. They wrote for the most part about their national architecture, two main periods of interest being precisely those relevant to St-Eustache: the Gothic and the newly defined field of the French Renaissance. Thus, virtually from the moment the structure/ornament St-Eustache was created, this identically configured building appeared in the texts of writers who comprehended it in quite different historical terms depending on whether, like Viollet-le-Duc, they understood the French Gothic to be the exemplary national mode of architecture or instead assigned this role to the French Renaissance, as did many of the Romantics. Furthermore, St-Eustache is a very large monument prominently located in the center of Paris, so that even when the French Renaissance (or, more typically, the church architecture of sixteenth-century France) has been the subject of little scholarly interest, it is a building that is difficult to ignore. Anyone writing about the history of Parisian architecture, or of Renaissance or classical architecture in France, or about the end of the Gothic has been more or less obliged to include St-Eustache, cumulatively providing ample material for my analysis.

I do not propose to begin by critically dismantling the structure/ornament pair, for although it is a truism that any attempt to describe an object will be a fundamentally interpretive and historically contingent act, by its very nature open to rigorous reexamination, at the same time another often overlooked factor needs to be considered. That is, any description of an object, in this case an architectural object, does more than demonstrate that a building has been seen in a particular historically specific way: it also produces a textual figuration or figure of the building. Such a figural building, whatever may be its relationship to the physical building it is seeking to represent, has its own discrete existence. It is a cultural artifact of value worth studying in its own right. A figural building is valuable in part because it has a certain utility. This utility is not restricted to the ability of the figure to convey knowledge of a building but is also specifically textual or literary in nature. Consequently, I temporarily want to leave the structure/ornament St-Eustache intact and begin by undertaking this more positive line of inquiry and consider what useful function this figural building might possess.

Structure/Ornament and the Critical History of Architecture

When I say that the modern structure/ornament St-Eustache is a figurative building I mean this literally. The binary structure of the concept is just that: a figurative structure or construction that is metaphorically composed of two closed and distinctly separate spaces. Onto the apparently neutral surfaces and into the apparently empty spaces of this figurative structure a variety of observations can be placed. Structure/ornament is a figuratively conceived heuristic device that provides architectural historians with spaces to be filled, a structure to be embellished.

One useful consequence is that the structure/ornament St-Eustache is a wonderfully genial building that makes itself available to a range of critical assessments. There is no predetermined correspondence between the structure/ornament building and what is said about it - what is placed in or on it - and the figured building confirms its own validity as it remains stable from the early nineteenth century onward, despite its shifting critical fortunes. For instance, Marius Vachon (1910) describes the church architecture of sixteenth-century France in a manner antithetical in judgment to Viollet-le-Duc's harsh characterization: "on a skeleton that is entirely Gothic, with traditional architectural schemas, they toss, in a charming caprice of the imagination, a Renaissance garment and adornment."(21) And he writes of St-Eustache:

As a whole, Saint-Eustache is a church with a skeleton of the Gothic type, overlaid with Renaissance adornment. Of the great cathedrals of the Middle Ages it has the boldness and majesty of construction; of the civic monuments of the sixteenth century it has the fantasy, the grace and the elegance of sculptural ornamentation. And nothing more luxurious, more delicate or more refined can be imagined.(22)

An imposing Gothic structure replaces Viollet-le-Duc's corrupted skeleton, elegant sculptural fantasy replaces Roman rags, and an architect of charming sensibility replaces Violletle-Duc's depraved rag and bone picker scavenging in the debris of the past.(23) Each author is in agreement about the essentially binary nature of the building and has so configured it, but in applying a different rhetorical veneer to its two separate parts is able to persuade us that St-Eustache is either a miserable or fine work of architecture.

Moreover, in each text the persuasive rhetorical veneer applied to the structure/ornament St-Eustache performs a narrative as well as critical function. The highly charged language that compellingly characterizes the church serves as a supplement to the narrative logic that organizes the story of the encounter of Gothic and Renaissance styles in sixteenth-century France. In the Dictionnaire Viollet-le-Duc tells a sad story of the decline, perversion, and eventual suppression of the French national mode as the seductive foreign forms of Roman architecture are insinuated into a weakened Gothic system. Vachon's La Renaissance francaise, to the contrary, narrates a positive encounter as a fertile medieval tradition nourishes and is in turn enriched by new architectural forms.(24) Each plot verifies the assessment of the structure/ornament St-Eustache; by the same token, the bipartite building affirms the validity of the plot as the descriptive terms that embellish it serve to sustain the historical story in which the building is a passing moment.

That the structure/ornament St-Eustache can accommodate (house and shelter) a variety of historicized scenarios is crucial to its tenacious success, for in modern architectural discourse narratives of the history of architecture - the historicity of the history of architecture - often constitute the ground for critical evaluation of architectural form. That is, in stating that the modern St-Eustache can accommodate different critical assessments, what I am really saying is that it can accommodate different dramatizations of the Gothic-meets-Renaissance story. These alternatives can include stories that contradict Viollet-le-Duc's, even those that also subject the building to a negative aesthetic appraisal.(25) Anthony Blunt, for example, writes:

It is to be expected that Gothic tendencies should survive longer in ecclesiastical architecture than in secular, and this is amply borne out by St Eustache. . . . This Gothic structure is, however, clothed in Renaissance forms. . . . the Italian impression depends only on the use of classical pilasters instead of Gothic. The orders are, it is true, used in a way to horrify any classically trained architect. In some piers, for instance, the four main faces are decorated with Corinthian pilasters, the height of which is perhaps twenty times their breadth, and the corners of the piers are filled by three columns standing one on top of the other, all of somewhat bastard design [ILLUSTRATION FOR FIGURE 2 OMITTED].(26)

The architectural hero of Blunt's Art and Architecture in France 1500-1700 is Viollet-le-Duc's despised antihero, the French classical ideal. Blunt charts the fortunes of classicism in a narrative of emergence, development, and triumph, in which the sixteenth century is a period of origin and progress rather than finale and decline. The Gothic, no longer cast in the role of Viollet-le-Duc's tragic victim, becomes an annoying, if historically expected malingerer unwilling to recognize that its time is up.(27)

All this is apparent in the description of St-Eustache, where superannuated Gothic "tendencies" are juxtaposed with badly proportioned and bastardized classical orders and where Blunt is easily detected in his textual persona of a horrified "classically trained architect." Blunt uses a historical narrative of progress to posit both the Gothic and St-Eustache as problematic, not because Gothic equals bad architecture, but because in the sixteenth century Gothic equals the past, what he calls "the old," and is therefore resistant to progress, that is, to the future, to "the new" of classicism. Resistance is, however, manifest also in the badly conceived and poorly executed Renaissance ornament of St-Eustache, which, nevertheless, announces the classical future and the ultimate futility of resistance.(28)

Structure/Ornament and Transitional Architecture

The self-chronicling of the structure/ornament St-Eustache, that is, the correspondence between the building on the one hand and the critical history of architecture on the other, is possible because just as every author sees and configures the same binary construction, so too is every narrative predicated on the evolutionary concept of a transitional period style: the building represents the passage of historical styles from the Gothic through the Renaissance and toward classicism. For each author St-Eustache is in transit, neither entirely departed from its Gothic origins nor fully arrived at its Renaissance or classical destination. As a result, Viollet-le-Duc is able to include the building in his story of Gothic architecture as persuasively as Blunt does in that of classical, and Vachon can narrow his sights on the transitional French Renaissance. There is no epistemological contradiction between the fact that they all see the same bipartite St-Eustache yet situate it in different dramas; the transitional Gothic structure/Renaissance ornament building itself seems to generate these alternate possibilities. It is all a matter of viewpoint provoked by the same historical phenomenon of the transitionality of the building, which can be subject to different dramatic spins and inflections.

There is a manifest correlation between the binary apprehension of St-Eustache as composed of a Gothic structure and Renaissance ornament and the historical comprehension of the building as a transitional one located at the end of the Gothic period and the beginning of the Renaissance. The evolutionary concept "transitional" presupposes continuous linear movement and narrative. Similarly, it must be recognized that the organization of the pair structure/ornament is not static or bidirectional but consistently sequential and thus inherently narrativized and endowed with historicity. "Structure" has temporal priority over "ornament," a status it enjoys both in the way the two words are normatively ordered (to speak of the ornament/structure concept would be deliberately perverse), and in the way architecture is itself conceived and built in the modern period. The (metaphorical) understanding of the history of architecture as being like a line in continuous forward motion dictates that Gothic moves transitionally toward (or through) the Renaissance, while the temporal organization of structure/ornament confirms the duality of perception on which the idea of transitional is dependent and the historical priority of Gothic, and also affirms the sequential motion from the Gothic to the Renaissance. Because of the homology between "structure/ornament" and "transitional," the material building as perceived by modern observers and as figured in their texts can function as an expressive synecdoche for the broad period of architectural history of which it is a transitional fragment.

Laugier had also used a spatial metaphor when he sought to describe St-Eustache as manifesting a style that was both classical and Gothic. He wanted to illustrate the duality of the building by making use of the spatial example of "bordering provinces," and he used space as a metaphor for time: the area where the languages and habits of two provinces overlap is like the moment when St-Eustache appears. If this metaphor is pursued its logic cannot be sustained. The space where neighboring provinces overlap is a nonspace with no internal integrity and no outer borders of its own; ultimately, it belongs more to one territory than another, or one territory will dominate and claim it. Indeed, a few traits from the classical province are the only ones that Laugier describes with any acuity, whereas Gothic is a nebulous, unspecified presence described as dying by the end of the passage. Finally, the spatial metaphor collapses; the borderland is diminished to a border, a line without space where no structures can be erected. When Laugier saw the building again he decisively centered it in the space labeled Gothic.

With structure/ornament, space is again used as a metaphor for time, except now the spatial metaphor is not a hypothetical place external to the building but is presented as a faithful description of the building itself which naturally motivates it. The moment when St-Eustache appears is like the figurative Gothic structure/Renaissance ornament building that transcribes its own transitional status. In other words, structure/ornament does more than accommodate historicized scenarios, it a priori obliges them; the inherent temporalization of the structure/ornament figure not only demands that the building be seen in strictly binary terms, it also functions as a compelling narrative device. The figurative pair permits architectural historians to be historians, allowing them to see and write about a building in such a way that it naturally conforms to and promotes their desire to tell continuous histories of the history of architecture. The primary textual utility of structure/ornament is that it is a figurative construction whose sequential organization narrativizes the observations that are placed on or in its two component spaces.

Furthermore, to avoid the structure/ornament St-Eustache is to avoid a historical understanding of the building, to ignore its transitional place in the history of architecture and the question of its period style. Such an evasion occurs in a description of St-Eustache by Viollet-le-Duc in the 1867 Paris Guide:

In the interior, the piers present the strangest profusion of pilasters and columns that it is possible to imagine. The effect of the whole of this interior, nevertheless, produces a seductive impression of elegant grandeur. Those elevated side aisles flood the nave with a beautiful, well-diffused light. There is certainly in all of this interior a theatrical affectation, the evident desire to astonish, and if this vessel was entirely painted, if the windows were furnished with lightly colored stained glass, the interior of the church of Saint-Eustache would have all the appearance of a fairy palace, if not of a Catholic church.(29)

Here, writing in the "nonserious" genre of the guidebook Viollet-le-Duc adopts a mode of discursive visuality different from that of the Dictionnaire. This decidedly nonanalytical, nonrigorous, and poetic mode allows him to offer a generally sympathetic response to the church by ignoring the question of its style, its place in the history of architecture, and its material division into structural and ornamental traits.

Conversely, to continue to rely on the structure/ornament figure will invariably serve to affirm the transitionality - or the inherent temporality - of a building even if an author seeks to avoid this term. Although such an escape has not been ventured for St-Eustache, Willibald Sauerlander has recently attempted to do so for a group of medieval Rhenish churches that since the early nineteenth century have been seen as transitionally situated between the Romanesque and Gothic. Sauerlainder contrasts the formalist transitional reading with what he construes as an alternate interpretation found in nineteenth-century texts, an option he calls the "ethnogeographic." Carl Schnaase is cited as an example:

Schnaase . . . tended to explain art by the influence of climate, soil and local customs. . . . He finds all sorts of features in the character of the Rhenish population and Rhenish landscape, which explain for him the decorative exuberance and the picturesque quality of the Rhenish transitional style. If one reads through Schnaase's pages one soon observes that while he keeps the word transition, in reality he sees the Rhenish monuments not as transitional but as creations of an autonomous regional style that leads not from Romanesque to Gothic but has aesthetic value in itself.(30)

This is a misreading of Schnaase in its presumption of an opposition between "autonomous regional style" and "transitional." Schnaase is not incorrect in claiming that he sees transitional architecture: he is deeply committed to this evolutionary stylistic description - for which he provides a contextual scenario. No epistemological inconsistency exists between the evolution of architecture, which describes these buildings as transitional, and contextualism, which examines the uniqueness - the autonomy - of their particular transitionality as the manifestation of a singular historical and cultural setting. Also, explanations in terms of national identity and specific regional and material requirements tended to be offered in the nineteenth century for transitional periods that negotiated simultaneously between a stylistic past and future as well as between local and foreign styles, whether German Romanesque and French Gothic or, as in sixteenth-century France, French Gothic and Italian Renaissance.(31)

Having distinguished between the two options, however, Sauerlander rejects both: "In my view neither of the two perspectives - the transitional and the ethno-geographic - is really satisfactory. But, we may ask, is there no other alternative? For the answer we need to take a fresh look at the monuments."(32) After a traditional formal analysis of a number of buildings Sauerlander poses what for him is the crucial question and offers a possible path toward its answer:

Why did Rhenish architecture only modernize decoration and not structure? I can't give an explanation but the answer given by those who refer simply to the German or Rhenish mentality is no more than self-adulatory. It would need a closer look into economic history, history of craftsmanship and technique, patronage and funding in order to come perhaps closer to an answer.(33)

The avenues of critical inquiry proposed here do not differ in kind from those pursued by Schnaase. Rather than interpreting these medieval churches in the context of regional identity, Sauerl/nder suggests a consideration of other contextual factors that might be called the "econo-technical" rather than "ethno-geographic." Like Schnaase, Sauerlander sees the buildings and their style, and he searches in an exterior context for an explanation of what he sees. Of greater interest than his explanation, however, is the question Sauerlander asks: "Why did Rhenish architecture only modernize decoration and not structure?" If the implications of this question are followed they lead in a direction the author certainly did not intend: Why are the structures old and the decoration new and progressive? Why does structure belong to the Romanesque past and ornament to modernity and the Gothic future? Why, in other words, are these buildings transitional?

In seeking to displace a concept that he sees as highly problematic, Sauerlander is inexorably drawn back to it and to the hypnotic linearity of the history of architecture as he employs an architectural description (also used by Schnaase) that perpetuates the status of these Rhenish buildings as in fact nothing other than transitional. His reliance on "structure/ornament" and his advocacy of the "econo-technical" rather than "transitional" and the "ethno-geographic" duplicate the interpretive strategies of Schnaase and do not move beyond them. While an inquiry into the "econo-technical" might be productive, it cannot be productive in the way Sauerlander hopes - that is, as capable of offering an interpretation of the buildings that would escape the conceptual structures of the nineteenth century - as long as these structures and their figurative guises are not themselves recognized and critically confronted.(34)

History and Architectural Metaphors

That architectural historians remain blind to the figurative nature of the structure/ornament pair is not only because the figure has effectively assumed the status of a true transcription of much historical architecture but also because since the early nineteenth century, history in its two dominant guises (the self-contained evolution of architectural history and contextualism) has often been posited as the singular and stable ground for architectural analysis and interpretation. Architecture is the object of knowledge; history is the mode knowledge takes to access this object. In the debates about what kind of historical or historicized narratives should be told about architecture, what has gone largely unnoticed is the way that architecture and the language of architectural discourse have provided figures to think about architecture itself.(35)

The preeminence accorded history by nineteenth-century theorists of architecture and by twentieth-century architectural historians is well known. We hardly need Michel Foucault to tell us, "History, from the nineteenth century, defines the birthplace of the empirical. . . . History has become the unavoidable element in our thought."(36) To study the practice of modern architectural discourse, particularly as it was originally formulated in the nineteenth century, is to encounter history and its metaphors at every turn as we are offered archaeology, history, historicity, continuity, progress, linearity, narrative, and temporality with a vengeance. Derailed histories of architecture are narrated; architecture is conceived as an object of historical knowledge and as an entity whose fundamental essence is historical (whether located in the diachronic movements of the history of architecture or synchronically correlated to "real" history); theories of architecture are now based on the study of history and on archaeology (that is, on a positivist, historicized empiricism rather than the idealized empiricism of earlier studies of historical architecture); these architectural theories are often written as histories, and architects conceive of their own work in historical terms?

But this aggressive foregrounding of history obscures the role that architectural and spatial metaphors play in much architectural discourse in both the nineteenth and twentieth centuries. A historian writing (in 1986) about French architectural theory of the nineteenth century states, "It was a discourse conducted in metaphorical terms; structured by the language of history and archeology."(38) This observation is partly right: architectural discourse of the 1800s and 1900s is conducted in metaphorical terms, but it is not - cannot be - structured by the language of history and archaeology. Properly speaking, "structure" belongs to architectural discourse; only as a metaphor can it make a transterritorial migration and "structure" historical language.(39)

It would be more accurate to say that the language, assumptions, methods, and theories of history and archaeology, the diachronic movements of the one and the synchronic probings of the other, are themselves structured, ordered, and conceived in terms of architectural and spatial figures, so that in a partly self-reflexive movement architecture is involved in the process that sees itself. The ascending and descending, forward or cyclical motions of the line of the history of architecture, for instance, cannot be charted unless there also exists a metaphorical structure or spatial matrix that allows that line to be plotted. In modern discourse, architecture and history are often mutually engaged in a nonseparable affirmation of the perception, conception, and description of architecture, whether historical, historicist, or historicized. Architectural metaphors (such as "structure") and architectural figures, such as structure/ornament, serve as spatial models that collaborate with historicized concepts, such as transitional. But a spatial model can be dangerous, for, as Jacques Derrida warns, "When the spatial model is hit upon, when it functions, critical reflection rests within it. In fact, and even if criticism does not admit this to be so."(40)

Structure/Ornament as a Narrative Device: A Closer Look

As a spat'lal model that also has the advantage of temporality, structure/ornament is fated to preserve the transitionality (or earliness or lateness) of buildings it configures, even if "criticism does not admit this to be so." By the same token, its temporalized bipartite structure permits and encourages different stories about a building to be written from those that concern the linear history of architecture. If a building is described as cleft between these two categories of traits, any observation about such a building will be located in one of its two component parts - its two figurative spaces - that are kept rigidly separate and always appear in the same chronological sequence, permitting and controlling narratives and ensuring that all observations are in fact narrativized.

To illustrate this more inclusive narrative or historicist role of the structure/ornament pair and to consider further its relationship to the transitional, I begin with a comparison of different accounts of St-Eustache that appear in Albert Lenoir and Leon Vaudoyer's critical history of French architecture, "Etudes d'architecture en France," which was serialized in the Magasin pittoresque between 1839 and 1852.(41) The first occurs in their discussion of the end of Gothic architecture, where the builders of St-Eustache are favorably contrasted with their contemporaries at the cathedral of Beauvais. The Parisians had the good sense to face the historical music and accept that Gothic had had its day, and therefore they made the appropriately enthusiastic effort to embrace the new Italian Renaissance forms [ILLUSTRATION FOR FIGURE 3 OMITTED]. Their colleagues at Beauvais, however, foolishly clung to the Gothic because they were "jealous of the success that Michelangelo obtained with the construction of the cupola of Saint-Peter's in Rome, and wanted to prove that the Gothic style could not only equal but surpass the great achievements of Greek and Roman architecture."(42) That their ambitious, staggeringly high crossing tower collapsed is noted by Lenoir and Vaudoyer as the hardly surprising result of such hubris.

The eager beavers at St-Eustache, meanwhile, were caught in a double bind:

In this church where the Gothic skeleton is preserved in its entirety, they wanted to apply the decorative elements which had been newly restored to honor; the round arch was substituted for the pointed arch in all the bays (with the exception of the apse) and the look of antique orders was introduced for all the supporting members; but was this attempt truly successful? And although at first glance this church offers a very seductive overall impression, are we not soon struck by the absence of harmony that must necessarily result from the application of these orders, the proportions of which are fixed by strict rules, to these immense Gothic piers, which are destined to support vaults whose skyward flight remained without limits? . . . Was it possible to introduce into such a complete creation . . . elements borrowed from an entirely differently constituted art? . . . We don't think so; and since the art of the West had to succumb to the influence of the Italian Renaissance, it is certainly in religious architecture that this may be regretted.(43)

In other words, the builders had to do what they had to do and if that resulted in bad architecture, well, that's progress for you: St-Eustache, trapped in a history beyond its control, was just as doomed to fail as Beauvais. Both the collapsed tower of the latter and the Gothic structure/Renaissance ornament St-Eustache serve the same evidentiary purpose, proving that in early sixteenth-century France the Gothic was finished and the formal dominance of the Renaissance inevitable, whether this future was resisted or met head-on.

What I want to draw attention to about this first account is that in it St-Eustache is entirely accommodated to the closed, unbroken story of the history of formal evolution in architecture; its facture is told exclusively in these monosystemic terms, and no social or contextual narrative is included. For reasons that I will soon consider, Lenoir and Vaudoyer do not call St-Eustache transitional. Yet their description of it conforms to the concept understood in its normative sense, and there is a perfect alignment between this evolutionary term and the structure/ornament building.

A very different scenario unfolds when Lenoir and Vaudoyer return to St-Eustache in their discussion of church architecture of the French Renaissance.(44) Here the continuous narrative of period styles completely breaks down. St-Eustache is now perceived as a drag on the movement of architecture because of the presence of its Gothic structure; moreover, all ecclesiastical construction of sixteenth-century France is so characterized. In the first place, we are told, there was very little church building in this period; second, much of what did get built was Gothic; and third, in those few instances where new Renaissance forms were taken up by a handful of enterprising architects, their efforts consisted of little more than the casual application of decorative features "accidentally thrown here and there" onto unchanged Gothic structures.(45)

This combination of Gothic structures and superficial Renaissance ornament was repeated again and again, from the beginning to the end of the century, in such works as the apse of St-Pierre in Caen, the transept facade of Ste-Clotilde in Le Grand-Andely, and the west facade of St-Michel in Dijon [ILLUSTRATION FOR FIGURES 4-6 OMITTED]. The unhappy result is that "in France we still do not possess a complete church of the sixteenth century conceived entirely according to Renaissance principles; St-Eustache was not even finished [until the seventeenth century], and, moreover, even there one sees but a church where the skeleton remained Gothic and which they wanted to dress in the fashion of the time."(46) In that one exceptional instance at St-Eustache when architects were given the opportunity to create not a piece of a building but a whole new important work in the Renaissance mode, the same solution is reproduced: a new Gothic "skeleton" is constructed and "dressed" in Renaissance decoration.

Others have been able to see a progressive (if often faltering and recalcitrant) classicism in the church architecture of sixteenth-century France, but Lenoir and Vaudoyer, who desperately want to see progress and development, change and continuity in the history of architecture, are stymied.(47) For them each effort represents a first step in the evolution from Gothic to a true Renaissance mode of church building, a first step that is endlessly reiterated. There is never any next step, never any follow-through. No developmental sequence can be imposed on this static series of identical and largely fragmentary gestures, and the history of sixteenthcentury church building shatters into a nonhistory, a story that should have happened but never got past page one.

I would argue that it is for this reason, rather than the described qualities of any given work, that both the individual projects and the period as a whole are viewed as problematic by Lenoir and Vaudoyer - and that they refuse to call St-Eustache and the church architecture of its time transitional. Unlike current scholars who reject "transitional" as outmoded and conceptually flawed, for nineteenth-century Romantic theorists it was a new and critically positive term used to describe periods of particular interest where both the evolutionary processes of architecture and the causal links between history and architecture were laid bare for analysis - and emulation.(48) Two such periods were the Italian Renaissance (by which was meant the trecento and quattrocento) and the French Renaissance minus its church construction. In fact, the "Etudes d'architecture" championed the secular architecture of the French Renaissance as the French national mode of building, whose principles, corrupted and then abandoned in the sterile architecture of seventeenth-century classicism, should now be revived to serve as a model for contemporary architecture.(49)

Lenoir and Vaudoyer write approvingly, for instance, of the early sixteenth-century chateau of Gaillon [ILLUSTRATION FOR FIGURE 7 OMITTED]:

In examining the fragments of this chateau . . . one sees that the style of its architecture was mixed. Next to the reproduction of the orders borrowed from antique art, certain details indicated that the Gothic influence was not yet without effect. . . . there is nothing shocking in the mingling of these styles, and this freedom of ornamentation created a very picturesque and very gracious impression. . . . the remains of this building are extremely precious for the history of art, and as complete models of this transitional period they may be profitably studied.(50)

Whereas the "mixed" style of St-Eustache is condemned for its lack of harmony, at Gaillon a similar phenomenon is perceived as charming and worthy of the most careful study. Read by itself the original lengthy description of St-Eustache cited above (to which the authors constantly refer in later discussions) might well suggest that Lenoir and Vaudoyer criticize St-Eustache because it represents, in the words of a scholar writing in the 1980s, "an unwitting or forced marriage of distinct and unrelated styles."(51) But their stand on French Renaissance secular architecture belies any such interpretation.

We can also consider their thoughts on early Italian Renaissance church construction as echoed in the words of their fellow Romantic advocate of transitional architecture, Leonce Reynaud:

These were still Gothic buildings, but with purer and more graceful forms, overlaid as it were by an alien veil, a veil rich and diaphanous, which decorated without concealing. There was a delightful blending of art and naivete in all this architecture, an exquisite taste and a great refinement. There was even originality, the borrowing from antiquity notwithstanding; for, if some details had been imitated, they had been brought together in a new way; there had been nothing servile in the copying, and especial care was taken not to alter in any way the general forms called for by the customs of the time.(52)

There is certainly room in this type of analysis, which speaks of "naivete" rather than "faults," to accommodate favorably St-Eustache as an individual monument. But this historicizing gaze does not see buildings individually; rather, it evaluates them according to their contribution to the continual progress of architecture. To resist this role, as St-Eustache and its contemporary church structures do in the texts of Reynaud and Lenoir and Vaudoyer, is to cause profound epistemological anxiety.(53)

In these texts what can be seen to distinguish the buildings of the Italian trecento (and quattrocento) and French Renaissance secular architecture from French ecclesiastical construction is that in the former cases the promise of progress signaled in each separate transitional monument is subsequently fulfilled, and the buildings can be strung together in a motivated linear sequence. In each work the future heraided is a future that arrives. Thus, in the "Etudes d'architecture," the transitional secular architecture of the reign of Louis XII is said to evolve into the less transitional architecture of the early years of the reign of Francis I, and eventually into a truly French Renaissance mode during the later years of Francis and the time of Henry II.(54) Of the early sixteenth century Lenoir and Vaudoyer wrote, "they still proceeded by trial and error, and as this is to be expected in transitional periods, matters of taste not yet having been stabilized, they indiscriminately mixed all the styles; the majority of buildings still preserved numerous traces of the Gothic style, which was only progressively abandoned."(55) It was precisely this progressive letting go of the Gothic that they were unable to plot in contemporary church architecture, which could not, therefore, be seen as properly and commendably transitional.

Furthermore (and contrary to the stated intentions of the authors), what makes the Italian and secular French works praiseworthy is that the promise of progress in each building is not necessarily or even primarily either a stylistic or social progress but a narrative one: point A always leads to point B, even if point B is only narratable as a decline (as happens when first in Italy and later in France the "transitional" Renaissance periods eventually give way to a rigid and doctrinaire classicism). It is this narrative role that Lenoir and Vaudoyer find themselves unable to assign St-Eustache in their discussion of the sixteenth century. The main problem posed by St-Eustache is not that it offends a theoretical position about style in architecture but that it thwarts the authors' efforts to move the story of architecture along: it does not "go" where they want it to; indeed, it does not go anywhere.

Yet if St-Eustache cannot contribute to the formal plot of the "Etudes d'architecture," the description of the building as stylistically divided between its Gothic structure and its Renaissance ornament indicates that despite Lenoir and Vaudoyer's refusal to call the building transitional (an evasion to which they certainly do not draw attention), they nevertheless fundamentally recognize the church as transitional. That their pro-transitional, pro-French Renaissance and anti-classical arguments could easily be adapted to St-Eustache was realized by many of their contemporaries. One mid-nineteenth-century admirer of the building, for example, asserts:

Renaissance style is the name commonly given to that transitional architecture where the pointed arch flattens and gives way to the round arch of the Greeks and Romans. Saint-Eustache is assuredly the most beautiful expression of this architecture, which was born in Italy at the end of the thirteenth century, but which soon lost its proper forms to servile imitation.(56)

Toward the end of the sixteenth century when, according to Lenoir and Vaudoyer, the principles of the French Renaissance were foolishly rejected by the classicists who mechanically imitated the architecture of antiquity, the narratives of French church and secular architecture reunite. Until that point is attained, however, they must camouflage the sizable gaps that occur in the history of church building as it breaks down into an inert chronicle of isolated and identical efforts scattered over the course of a century. They also feel obliged to explain why French church architecture could not emulate the route traced by Italian church architecture, which negotiated so successfully between the medieval and the revived antique (and which led rather than followed developments in secular building), and above all, why it did not keep pace with the admirable, socially responsive efforts apparent in transitional French chateau and civic construction.

Lenoir and Vaudoyer now have little choice but to turn to contextualism, which offers them a multitude of explanations: France was already blanketed with churches and the demand for new ones was consequently low; unlike the Italians, who had a strong national affinity for the antique, which allowed them to make a clean break with the Gothic, antiquity had no comparable meaning for the French; more important, because of the rise of Protestantism the sixteenth century was a period of religious crisis, "of wars and endless massacres," and the French clergy had other things to worry about besides the commissioning of new churches, nor were they willing to risk weakening the position of the Catholic Church by encouraging substantial changes in their architecture; unlike Italy, where the Renaissance of architecture corresponded to a religious renewal, no parallel phenomenon occurred in France, where the new style primarily satisfied material rather than spiritual needs and found itself most vigorously developed in residential architecture.(57) In their words: "The Renaissance of French architecture . . . was a protest of sensual inclinations against the mortification imposed by Christianity and against the rigorous austerity of medieval mores."(58)

Lenoir and Vaudoyer were great advocates of the view that architecture and history were involved in a dialectical process:

The particular character that distinguishes each of the major periods of history can be easily determined by that of the art that corresponds to them, and reciprocally, the successive transformations of art can only be truly appreciated when we link them to the social principles of which they are the consequence.(59)

Church architecture of the French Renaissance, however, defeats their desire to chart the reciprocal relationship between architectural and historical forces, and they are driven to rely exclusively on the latter to describe the nonprogress of church building in this period; contextualism here assumes the function of a dissembling prosthesis that allows a broken formal history to be made apparently whole.

In the "Etudes d'architecture" the structure/ornament St-Eustache appears in two very different stories: the first is effortlessly told in terms of the continuous evolution of period styles; in the second this seamless narrative falls apart and recourse to historical explanation and contextual narrative is made urgently necessary. The structure/ornament figure houses not only two different narratives but also two fundamentally distinct, even contrary, lines of argument concerning the facture and historical meaning of StEustache. According to one, the presence of both the Gothic structure and Renaissance ornament is unproblematic and entirely comprehensible as the building embodies the passage from the Gothic past to the Renaissance future. According to the other, the Gothic structure signifies the resistance, inherent conservatism, and lack of will on the part of the clergy in sixteenth-century France, while the Renaissance ornament, no longer interpreted as a laudable desire to obey the historical mandate of progress, becomes a banal attempt to evoke the current architectural fashion, to be "a la mode du temps." In each case the phenomenologically stable building is subject to the identical aesthetic evaluation, but the terms of inquiry belong to two different modes of analysis, each of which is, however, coordinated to and controlled by the sequential narrative structure of the structure/ornament pair.

Each narrative also grants a different figurative status to St-Eustache. When it is transitionally located in the evolution of architecture it becomes a synecdoche designating the historical totality of which it is a part. As a synecdoche St-Eustache is indissoluble from the historical panorama it represents, whereas in the contextual narrative a gap appears between the building and its historical context; St-Eustache now becomes an extended metaphor or allegory of the troubled and conflicted times that produced it.

Structure/Ornament and the Multiple Narrative

In the "Etudes d'architecture" the two narratives about St-Eustache, one of formal evolution, one concerning social, religious, and cultural history, are kept apart in separate essays (appearing, in fact, in different issues of the Magasin pittoresque). But other texts use the sequential, binary organization of structure/ornament to graft together multiple discourses and narratives about the building, allowing this figurative structure to blur contradictions and permitting a variety of visual and critical viewpoints to be offered simultaneously. Such a use of structure/ornament can be seen if we reexamine Anthony Blunt's account of the building.(60) When cited above, Blunt's text was abbreviated in order to foreground his primary narrative of stylistic progress; a closer and more comprehensive consideration reveals that his reading of the building is far from homogeneous.

Blunt opens with a laconically cryptic historical interpretation of why church architecture in sixteenth-century France did not capitulate as quickly to the new Renaissance mode as did secular: "It is to be expected that Gothic tendencies should survive longer in ecclesiastical architecture than in secular and this is amply borne out by St Eustache." The stress on surviving Gothic "tendencies" speaks to the historicist ideal of continuous progress in architecture, which is here thwarted (by unnamed historical contingencies), and also to the passivity of a building apparently unable to control its destiny within that progress.(61)

He then continues, "It represents, however, a remarkable compromise between new and old, quite different from St Pierre at Caen. Here the plan, structure and proportions are nearer to High Gothic than Flamboyant. The plan is almost exactly that of Notre-Dame. . . . The proportions of the nave again recall the thirteenth rather than the fifteenth century." Having opened with a story of stylistic survival, Blunt shifts to a tale of architectural revival, as the designers of the building turn away from the most recent phase of Gothic (that is, away from the Flamboyant style that "survives" into the sixteenth century in works like the apse of Caen) and instead evoke the cathedral architecture of three hundred years previous. In addition, a specific source, Notre-Dame, is posited for the plan of the church. If Blunt first interprets the Gothicness of St-Eustache as the unconscious persistence of a past mode of architecture, he then sketches a wholly different scenario where the intertwined stories of the forward movement of architecture and its (enigmatic) historical context are now rejected in favor of the more circumscribed story of the design process of the building, where conscious intent and the deliberate use of sources necessarily come into play. What allows his conflicting interpretations to cohere is that they are placed in the same figurative space and are thereby provided with a counterfeit kinship, a likeness that is in name - Gothic structure - only.

Blunt draws no conclusions from his observations, nor does he call attention to his contradictory survival-yet-revival schema. Instead, he continues, "This Gothic structure is, however, clothed in Renaissance forms." This phrase, articulating the transitionality of the church, and which seems to describe its actual fabrication, also forms a textual transition as Blunt moves from a contemplation of the structural to the ornamental traits of the building, summarizing (and simplifying) what has been said - it all boils down to a generic, uniformly observed "Gothic structure" - and introducing what is to come:

The ornament . . . is very simple, and the Italian impression depends only on the use of classical pilasters instead of Gothic. The orders are, it is true, used in a way to horrify any classically trained architect. In some piers, for instance, the four main faces are decorated with Corinthian pilasters, the height of which is perhaps twenty times their breadth, and the corners of the piers are filled by three columns standing one on top of the other, all of somewhat bastard design.

Whereas the structural traits were simply observed and their possible sources identified, Blunt submits the ornamental forms to a crit'lcal aesthetic appraisal based on a normative ahistorical ideal of classicism.(62) Taken as a whole it can be seen that Blunt's text interprets each "part" of the building as resistant and aberrant, but for different reasons: the Gothic structure for simply being there, a historical hangover out of its proper place, the Renaissance ornament for violating the classical ideal. That he evaluates the two "halves" of the building according to two essentially unrelated modes of analysis - one concerned with the ideal of formal progress and the specific historicity of the building, the other with transcendental aesthetic ideals of classicism untrammeled by specific historic considerations - is masked by the temporal organization of the structure/ornament figure, which imposes a narrative (and historicized) coherence on them. Also, even if Blunt does not explicitly relate his observations about the classical elements to the ideal of formal progress, the structure of structure/ornament conveniently and implicitly accomplishes that for him. The building is thereby allowed to conform to his overarching narrative of the gradual emergence and final triumph of French classicism in the seventeenth century.

But Blunt is not entirely satisfied; he continues:

And yet, in spite of these eccentricities, the interior of St Eustache has a grandeur of space and proportions not to be found in any other sixteenth-century church in France. It is true that in these features it follows a medieval rather than contemporary tradition, and it must also be noticed that the church was to have no influence on the general evolution of French architecture; but as an isolated work it remains of great importance.

This passage begins with a return to a consideration of the Gothic structure of the building. That Blunt is primarily looking at this aspect is clear, for what he praises - the singular "grandeur of space and proportions" of the building - is attributed to "a medieval rather than contemporary tradition." What he means by "contemporary tradition," however, is confusing, for according to him, "church architecture during this period was in the main limited to additions and alterations to existing buildings." All these projects are described as fundamentally medievalizing, the Renaissance presence restricted to ornamental forms and the superficial use of the classical orders.(63) By "contemporary tradition," I would suggest, Blunt does not mean the actual historical reality of ecclesiastical building during this period but rather what this reality should have been, that is, less medieval and more devoted to the development of a classical mode of French church architecture.

Such a critique is presented in the concluding phrase: "as an isolated work, it remains of great importance." Isolated? From what? Here Blunt leaves no room for doubt: isolated from the history of architecture, from a consideration of the building's contribution to stylistic progress, from the recognition of its failure to abandon the medieval tradition and its inability to influence "the general evolution of French architecture." If we can forget that the building's ties to the past are too strong and its impact on the future nil, if we can momentarily suspend our serious scholarly faculties and suppress that St-Eustache is an evolutionary aberration, if we can blur our vision and see only its "grandeur of space and proportions," then we can conclude it is a work "of great importance."

In the final remarks the text flickers between two perspectives: one looks with anatomizing scholarly precision at the entire structure/ornament building and sites it in the evolution of the history of architecture; the other, which is barely allowed to function, looks through squinted eyes at the generally "medieval" (rather than specifically "Gothic") structural aspects of the building alone yet interprets these historically qualified traits in an isolated, dehistoricized context. This stepping out of history to view St-Eustache through a nonanalytical perspective is of course similar to what Viollet-le-Duc does in his description of the church as a "fairy palace." Such a self-consciously nonrigorous, ahistorical response occasionally appears in modern literature on the building, most often, as with Blunt, offered as an "on the other hand" view subordinate to the serious critique of the structure/ornament transitional building? Even Lenoir and Vaudoyer preface their critical analysis with the disclaimer that "at first glance this church offers a very seductive overall impression." These intermittent glimpses of a grand, seductive, and sometimes unreal church emerge from a discourse that might be called "irrational," for it appears as an inferior alternative to the dominant "objective" discourse by which the building is "rationally" seen and permits authors to articulate an affective response to aspects of St-Eustache that escape hard critical reflection - what they notice before they are compelled to move beyond that "first glance" and seriously gaze at the monument.

If Blunt's reading of the building as a whole is governed by the story of continuity and progress in the history of architecture, at the same time he splices into this controlling master narrative a number of other narratives and modes of analysis. In this action he is aided by the figurative structure of the structure/ornament pair that remains intact throughout, allowing the transitionality of the building to dominate, while also providing a framework for his multiple interpretative modes and viewpoints. This temporalized binary structure imposes a de facto narrative coherence on his heterogeneous observations (which are sometimes presented as little more than fragments, as in "It is to be expected . . ."). But if he is aided, he is also coerced. The diachronic structure of structure/ornament, with its two adjacent spaces that are sealed off from each other and that always appear in the same temporal order, both permits and insists that St-Eustache be seen separately and sequentially in terms of an uncompromising dualism, and it both allows and compels narrativized interpretations of the building.

In modern discourse, structure/ornament generally figures architectural form in such a way that a building synecdochically, metaphorically, or allegorically narrates and displays its own design process and fabrication, its location in the linear movements of the history of architecture, and its contextual motives and meaning. To varying degrees these possibilities are all realized in Blunt's exceedingly complex and nuanced, if erratic and contradictory, account of St-Eustache.

Looking for the Structure/Ornament St-Eustache

That the structure/ornament St-Eustache has dominated serious looking at the building since about 1830 is no testament to the veracity of this construction; rather, it pointedly underscores that the structure/ornament figure was and has remained the common point of departure for analysis, not its conclusion. It exists prior to research and looking, predetermining the shape that visual and discursive responses to the building will take. This a priori spatial model is a powerful one: by fracturing the building into two parts that cannot overlap, it permits no observation that bears on a structural element to seep into the space called ornament. Once the structure of St-Eustache is called Gothic (or medieval, or French), nothing ornamental can be claimed by that term; once its ornament is called classical (or Renaissance or Italian), all structure is removed from that domain. The implications of this cleavage multiply as the definition of each space is refined. Once a perceived desire for Renaissance modernity on the one hand or for historicist evocation on the other has been identified at St-Eustache, neither can inform both the Gothic structure and the Renaissance ornament of the building. Once its ornament has been called classical, no structural trait can be so named and Gothic itself must be understood as an architecture that is completely evacuated of all possible classicism.

If historians are unaware of the figurative nature of structure/ornament, it cannot be said that this figure is suppressed and hidden away. Instead, like the purloined letter, it is concealed through the simplest of camouflages: hidden in plain sight, it is openly displayed in the guise of a literal architectural description that masquerades as a transcription of what is really there. Structure/ornament allows the historicized gaze of modern viewers to see a building in historicized terms, at the same time that they are encouraged to forget the status of the pair as an a priori figurative construction. Modern architectural discourse would have us believe that its historical inquiries, its discovery of the historicity of architecture, of the phenomenon of transitionality, or of the historical forces that shape architecture are fundamental to the clarified view of St-Eustache as composed of two categories of morphologically identifiable period style traits. Again, history is foregrounded as the foundation for architectural knowledge, and a figurative structure, which is equally implicated in this process, is denied complicity by its presentation as a self-evidently literal description.

At this point, we may well ask: How literal is the structure/ornament St-Eustache meant to be? Certainly there is nothing in any of the texts cited to suggest that their authors do not mean what they say. But what do they mean? The material structure/ornament building is in fact a very difficult one to pin down. As soon as we try to take it at face value and seriously examine it, its apparent simplicity and clarity fall away; we find ourselves mired in contradictions and false starts, following avenues of analysis that turn back on themselves and discovering that apparently straightforward terms are in need of qualification.

Problems begin before the church itself is looked at, when we try to define the terms structure and ornament as descriptive of material architecture. In a strictly literal sense, that is, in architectural discourse, when we say that a building has a "structure," this noun signifies the material realization of the tectonic principle by which load, support, and thrust are accommodated, as distinct from the rest of the building. On the other hand, the word structure in English (as opposed to French) still retains one of the original Latin meanings of structura that allows it to denote literally the entire structure, the complete work of architecture itself, and it is a word used interchangeably with building, edifice, monument, and so on.(65) In this sense, structure includes the system of statics indicated by the more strictly tectonic meaning of the word, and it also encompasses the building's ornament. Structure thus describes a self-sufficient entity, the building as a closed object, a unified and autonomous presence.

Ornament possesses a different lexical status, for it does not belong first and foremost to the language of architecture. Derived from the Latin ornamentum, ornament is "anythingthat decorates or adorns; an embellishment. . . . a group of notes that embellishes a melody, . . . . one whose qualities adorn or confer luster on those about him," "any adjunct or accessory . . . equipment, furniture, attire, trappings."(66) Ornament can be as insubstantial as a rhetorical flourish or as weighty as the portico, temple, and colossus that Leon Battista Alberti tells us are the proper ornaments of a harbor.(67) Lexically parasitic, ornament only moves from the realm of the general and conceptual to the specific and physical when we know what it is applied to or where it appears.

In modern architectural discourse this lexical dependency mirrors the relationship between built structure and built ornament. Just as ornament depends on structure for definition in the structure/ornament pair, so too is the architectural object "ornament" seen to need built structure, without which it is but a fragment - whether a literal fragment, as are found at ruins and building sites, or images of fragments, as are found in books on architectural ornament (which proliferated in the nineteenth and early twentieth centuries).(68) Detached from its setting, ornament is a relic, a fetish, a sculptural souvenir - indeed, as Viollet-le-Duc sought to demonstrate, a memory of a whole object. Architectural ornament cannot become meaningful and achieve a wholesome, nonfragmentary presence unless it appears on a structure to which it is an adjunct or an accessory, secondary and contingent.

When considering the diachronic narrativized structure of structure/ornament the fiction could be maintained that we were dealing with two coequal entities. If ornament is literally placed in second place, its lesser status nevertheless seemed incidental; its function was equivalent to that of structure as it provided an equally spacious area for "half" of the building and stories about it to be located. But the pair is more than simply a binary one predicated on the mutual exclusivity of the two terms: it is also an implacably hierarchical one where ornament appears as a lexically and materially dependent feature, as an entity that is incomplete and inessential in and by itself.

At the same time, however, whereas ornament seems to achieve a stable meaning within the structure/ornament pair, the same cannot be said for structure. By itself structure may define an autonomous architectural presence, and does so whether it refers to the physical structural system of a building or to the building as a totality. But a third and far less secure meaning attaches to structure when it is paired with ornament, that is, when it occurs in the context in which we often encounter it in modern architectural literature. Here its meaning is modified to designate everything in the building except ornament. It is the building without its ornament or, given the temporalization of the pair, before its ornament is added, a process that must be reversed in the eye of the viewer. In this pairing structure emerges as a negative entity: it is not ornament; it is what is left over after ornament has been seen and detached. Structure becomes knowable not as an immediately apprehensible self-present presence, but as the residue of ornament, as the absence of ornament, as that part of the building that cannot make itself seeable to the viewer in the first place.

In the structure/ornament pair, structure takes on the qualities normally associated with ornament: it is the remainder, the needy secondary entity whose need - to be seen, to be known, to be present - is supplied by ornament. Structure does not display ornament; rather, ornament reveals and makes present structure and it does so by pointing to and compensating for what structure lacks. It completes structure, but not in the way normally thought: it does so by allowing structure to become whole and present before ornament is put in place; at the same time, it is only when ornament appears on the scene that this (pre-ornamental) wholeness is achieved.

Thus, ornament in the structure/ornament pair functions according to Derrida's logic of the supplement, in particular of the supplement that is the parergon (that which is next to, attached to, outside of the ergon, that is, the work, of art).(69) What constitutes parerga

is not simply their exteriority as surplus, it is the internal structural link which rivets them to the lack in the interior of the ergon. And this lack would be constitutive of the very unity of the ergon. Without this lack, the ergon would have no need of a parergon. The ergon's lack is the lack of a parergon . . . which nevertheless remains exterior to it.(70)

Ornament, the marginal and additional entity that is attached to, placed on the outside of, and comes after structure, turns out to be central, primary, and essential. Like a parergon, ornament remains exterior to structure, but its exteriority no longer signifies that it is easy to detach, or even that it is detachable at all.

And like the supplemental parergon attached to the ergon, it is by no means a simple or self-evident process to decide where ornament ends and structure begins. If the logic of structure/ornament demands that we start by removing the ornament, just how is this visual vandalization of St-Eustache to be done - and is it even doable [ILLUSTRATION FOR FIGURE 8 OMITTED]? To begin with, it is tricky to decide what is ornament and what is not, and we are further constrained by the fact that ornament is qualified by a stylistic label. We have to be careful to leave in place anything that might be Gothic, for the logic of the binary pair categorically insists that it belongs to structure. What do we remove and what do we leave alone? Although the guideline that we limit ourselves to Renaissance traits seems helpful, in fact it adds another level of uncertainty to our violent surgery, paralyzing our dissecting tool rather than clarifying its object. Would we really want to claim that the sculptural detail illustrated in Figure 9 depicts a Renaissance man?

The process is further complicated by the alternate methods by which the ornament of St-Eustache is said to appear: it is applied to the structure, it clothes the structure, Gothic ornament is substituted by or is translated into Renaissance ornament. The application and clothing methods are the most frequently suggested, but the substitution and translation procedures, which presuppose an altogether different degree of completeness on the part of the underlying original pre-ornament Gothic structure, are also offered. Paul Frankl, for instance, saw St-Eustache as an example of "the passive transition to the Renaissance" where one by one Gothic members are replaced by Renaissance forms.(71) And in 1984 Michael Hesse stated that" the whole church was submitted to the principle according to which Renaissance decoration in the prevailing taste was applied to the Gothic structure, each minute Gothic motif was translated into an antique element."(72) Although Hesse presents his "translation" remark as a corroborative amplification of his "application" observation, in fact he describes two mutually exclusive procedures. Or does he mean that some ornamental elements were translated while others were applied? And if the peeling away of the latter is fraught with difficulties, don't we truly lose our footing with the translated ornament? At which point do we stop and decide that it has reverted back to its original Gothic form? Also, can we be sure about the origin of the translated term and the direction the translation is taking? Is it possible that a Renaissance element has been translated into a more Gothic form rather than the other way around [ILLUSTRATION FOR FIGURE 10 OMITTED]? And even if we could arrest the retranslation with any degree of certitude, would we then not have a Gothic structure with Gothic ornament rather than a Gothic structure pure and simple? Furthermore, a process of translation is predicated on a real or established equivalence between two languages, or, in this case, two repertoires of architectural forms. But this necessary equivalence is precluded by the binary opposition of Gothic and Renaissance architecture, which rigorously insists on the absolute absence of any common ground between them.

Similar doubts concerning the ability to arrive at a knowable Gothic structure attend the clothing and application methods. If the structure has been "clothed" in ornament, what was the point of departure: a nude or denuded Gothic structure? Was the structure a Gothic structure before its ornament was added or a Gothic structure stripped of ornament and subsequently reclothed? Which meaning of the word structure attends "Gothic structure"? The matter is complicated by the fact that Gothic architecture - the Gothic "style" in architecture - is often understood to be consubstantial with its structural system. One architectural dictionary, for example, defines Gothic architecture as "the architecture of the pointed arch, the rib vault, the flying buttress."(73) If what is Gothic about St-Eustache is reducible to this system - as it seems to be when the word "skeleton" describes the Gothic part of St-Eustache - is it then the vast remainder that must be understood as Renaissance, as the ornament of the building?

All our best efforts to detach the ornament, to reconceive and reverse the process by which it was conceived and put in place, would be permanently stalled by the inability to agree on what ornament consists of or how it got there. And even if this process was feasible we would never arrive with any security at a stable structural presence, for it is unclear what "Gothic structure" means. Is the Gothic structure of St-Eustache a Gothic structural system, an entire Gothic building, a Gothic building with or without ornament, a Gothic structure before or after Gothic ornament was attached or taken away? We cannot know the structure of the building until the ornament has been removed, but to remove it we need to know what the structure was before ornament was added.

The description of St-Eustache as a Gothic structure with Renaissance ornament seems a simple one, but it is impossible to see. Structure is the recessive, unrecuperable, unstable presence that finally we cannot work hack to; ornament becomes all that we can clearly see, but we can never remove it, see past it, without destroying the structure, which is essential. They are both there but do not coexist in the simple oppositional way that so many modern texts would have us believe.

Structure/Ornament as Metaphor

This analysis might appear to be engaged in gratuitous nitpicking, which deceptively mystifies the issue by lighting on marginal weaknesses in the structure/ornament description of St-Eustache. It might be argued that what is plainly intended when St-Eustache is described as a Gothic structure with Renaissance ornament is that the building is essentially Gothic, that it adheres in all vital aspects to Gothic principles and underlying systems of organization; that the rest - ornament - is not Gothic, or maybe is a little Gothic here and there, but not in any essential or fundamental way; and that my insistence on literal accountability is, to say the least, naive.

Against this argument I offer two counterarguments. First, once we let "structure" signify what is "essential" and "fundamental," what concerns "principles" and "underlying systems," we have moved from the domain of the physical, visual, and literally architectural to that of the abstract and metaphorical, a move that is concealed by the apparent aptness of using structure/ornament - structure in particular - to describe the real material presence of architecture.

Non-architectural discourses have long appropriated the language of built architecture for metaphorical use. Earlier I cited one example of this process as structure was commandeered to the language of history and archaeology. Such borrowing of architectural terms has been fundamental to the enterprise of Western philosophy, but it also occurs throughout all scholarly discourses and in everyday language, where buttressed ideas, foundations of societies, keystones of arguments, pillars of the community, scaffoldings of political platforms, and other such metaphors regularly appear.(74) As Denis Hollier has written:

There is . . . no way to describe a system without resorting to the vocabulary of architecture. . . . nothing becomes legible unless it is submitted to the architectural grid. . . . It is as if, by allowing themselves to be named metaphorically by a vocabulary borrowed from architecture, the various fields of ideological production uncovered a unitary vocation. . . . Without architecture the world would remain illegible.(75)

What has not been sufficiently recognized, however, is that having been recast into metaphor the language of architecture can return to architectural discourse in its altered figurative state. Once general discourse has changed the meaning of architectural terms, they can be - have been - reappropriated back to architectural discourse in their new figurative guises in which they no longer possess a precise architectural meaning - indeed, they have taken on nonarchitectural significance. One relatively simple example of such reappropriation might be the statement: "Brunelleschi's study of Roman monuments formed the foundation of his late architecture." That "foundation" is used figuratively here is not in question; no one would confuse the actual masonry substructure of S. Spirito with Brunelleschi's Pantheon-enhanced knowledge. When structure is used to designate what is Gothic about St-Eustache, however, the figurative meaning does not unequivocally replace but is commingled with the literal. That structure has been away on a transformative voyage and returned home in its changed figurative condition goes unrecognized.

When discussing structure/ornament as a figurative construction I was not overly concerned with the semantics of structure, but its (overlooked) metaphorical meaning turns out to be crucial. In the typical binary description of the building structure often (covertly) signifies in the way that it does in the following sentence describing a collaborative mode of cyberspace fiction known as the Hypertext Hotel: "Anyone with Internet access can visit the hotel, and once there can embellish the existing structure (with marginalia)."(76) In neither case does structure designate the literal, tectonic, visible structure of the work (indeed, the Hypertext Hotel does not possess such a structure); rather, structure denotes the governing concept and essential idea - the plot of the hotel, the design of the building - which informs and organizes (in fact, structures) everything else. In the case of St-Eustache, this "everything else" comprises its material ornament as well as its material structure. Both literal structure and ornament - the entire physical building, in other words - metaphorically ornament the metaphorical structure of St-Eustache, its design.

Thus, the figurative signification of structure does not escape but instead confirms the parergonal and supplemental status of ornament. If structure in its metaphorical sense is the fully conceived structured idea of the building, then ornament becomes the parergon that reifies the underlying essential idea, the recessive unseen structure. Without ornament the structure would not be realized, would not be present. Consequently, when structure takes on its metaphorical meaning ornament does not follow it into the realm of the invisible but gains sensationally in visual and physical presence. The entire built building becomes the ornament, the necessary supplement that is added to and replaces its own structure, making the essential structure knowable.

I am not claiming, however, that the structure/ornament description of St-Eustache is only a metaphor, that the way it is used in modern texts on the building is really metaphorical and not at all literal. This is not where the epistemological blindness lies. Modern observers of St-Eustache do mean what they say, but they are not cognizant of all that they are actually saying, being unaware of both the metaphorical meaning of structure/ornament as well as the parergonal logic that inhabits both its proper and figurative meanings. The persuasiveness of structure/ornament in modern discourse depends on both the forgetting of its metaphorical meaning and the foregrounding of its literal meaning in simple binary terms. If viewers of the building start to experience a slippage in the literal description of the building as divided between its Gothic structure and Renaissance ornament the other metaphorical sense of the words is called into play. But in order to preserve the status of the pair as a literal objective description, its metaphorical meaning is called into play before such slippage actually occurs to the viewer: it is a maneuver that occurs within the epistemology and logic of the gaze, not the consciousness of the viewer. This unrecognized move has much to do with the continued success of the term as a description of the building. And, of course, the third function of the pair - as a spatial model that permits a range of narratives about the building to be told - will also continue to confirm the binary perception of the building (a perception that is further sustained by the rigid opposition of Gothic and classical).

These three functions of structure/ornament (as a literal description of the physical building, as a metaphorical understanding of its presence, and as a narrativized figurative tool) are all operative and intertwined in modern accounts of St-Eustache. Yet the only openly conceded role of structure/ornament is that of a strictly literal architectural description. The other two functions work silently behind and serve to support the facade of the literal meaning, and they do so with the unintentional complicity of those who use it that way. Or at least this is plainly the case in twentieth-century writings on the building, which brings me to my second argument against the view that structure/ornament is a perfectly adequate and common-sense description of the material St-Eustache.

Structure/Ornament as Metaphor in Nineteenth-Century Texts

Although the structure/ornament St-Eustache was constructed in the early nineteenth century, it is only beginning in the twentieth that this building becomes ubiquitous and overwhelms the scholarly St-Eustache literature, as was documented in the list of examples cited earlier. Unlike their more recent academic counterparts, French architectural theorists of the nineteenth century occasionally offered serious descriptions of the building that do not reiterate the structure/ornament paradigm. That is, different textual figurations of St-Eustache were produced not only by nonserious looking at the building (as when Viollet-le-Duc described it as a "fairy palace") but also by attempts to see and describe it in a scrupulous, historicized way. These otherwise constructed St-Eustaches are, nevertheless, in conceptual, theoretical, and metaphorical alignment with the structure/ornament building. If visually, as a physical object phenomenologically experienced and textually figured by the gaze, Saint-Eustache is not literally described as a Gothic structure with Renaissance ornament, it is, however, theoretically and metaphorically understood as divided between a Gothic "structure" and supplemental Renaissance "ornament."

Viollet-le-Duc, for example, eight years before he published the Dictionnaire description of St-Eustache, offered a different perception of the building in the periodical Annales Archeologiques:

Saint-Eustache is a badly built monument of the thirteenth century, which shocks by its lack of unity. Those side aisles of a useless height, those piers formed of an amalgam of pilasters and columns which are irrationally entangled, those vaults whose ribs are interlaced every which way and are no longer indicative of the real construction, those keystones fastened to the framework, those windows of a disagreeable proportion which seem to be squeezed in above that little triforium that seems more like a balustrade than a gallery, those mullions whose soft forms indicate neither stone nor wood construction, those flying buttresses with concave extrados, all these senseless combinations . . . are they a progress? In this case, does the antique element add anything to the beautiful arrangement of the scheme which is from the thirteenth or fourteenth century? We don't believe so.(77)

Viollet-le-Duc here presents an ekphratic antipanegyric where the detailed list of faults figures St-Eustache as a distorted thirteenth-century Gothic building, not as a Gothic structure dressed in Roman or Renaissance ornament (as it is in the Dictionnaire and in the texts of Vachon, Blunt, and Lenoir and Vaudoyer). The entire structure, the "monument," is Gothic, and the "antique element" is not granted an effable or visible material presence. It is first mentioned at the very end of the passage, where it is implicated as the cause of the debauchment of this "monument of the thirteenth century," of its aisles that are too high, its too small triforium, weirdly shaped flyers, and so on. Every misshapen feature here catalogued has its healthy counterpart in Viollet-le-Duc's description of a genuine thirteenth-century church that is configured as an organic, complete whole, its every feature coordinated to and expressive of a single unified principle.(78) This is what St-Eustache should have been, would have been, were it not for the insidious (yet apparently invisible) addition of the antique element, which here does not function as a good and necessary material supplement that completes and makes present the structure of the building. Instead, it becomes, in Derrida's terms, an abnormal parergon and "dangerous supplement" that "harms the beauty of the work . . . does it wrong and causes it detriment," "cuts into an energy which must (should) have been and remained intact. . . . that enfeebles . . . and falsities."(79) The antique element is an unnecessary corrosive extra whose intangible presence cruelly infects and perverts the fundamental ontological essence - the structure - of St-Eustache and produces this tragic dopple-ganger of a thirteenth-century church.

If the St-Eustache constructed in this passage is visually and physically different than the building that will soon materialize in the pages of the Dictionnaire, nevertheless, in each case the building is theoretically comprehended in the same historicized and metaphorical binary terms: it is essentially and "structurally" Gothic and only secondarily classical. In the Annales Archeologiques, however, the antique element is not so much literally exterior to the underlying Gothic structure as it is metaphorically comprehended as ontologically exterior to and alien to the essential reality of the building. It is this metaphorical, theoretically based comprehension of St-Eustache, rather than a defined perceptual apprehension, that informs Viollet-le-Duc's quasi-literal description of the building as a "badly built monument of the thirteenth century" in the Annales Archeologiques essay and as a Gothic skeleton freakishly outfitted in Roman tatters in the Dictionnaire.(80)

A more overtly figurative, that is, less literally architectural, characterization of the building had appeared the previous year in the Annales Archeologiques, written by Viollet-le-Duc's fellow Gothic rationalist and restorer Jean-Baptiste Lassus. Lassus describes St-Eustache as a building "where the Gothic principle exists in its entirety, but is completely denatured, weakened, and falsified by an envelope that is utterly Mien to it." This St-Eustache whose "principle" is perverted and masked by a strange "envelope" is extremely evocative yet is the least seeable of the modern St-Eustaches so far considered. The essence of the building is its physically insubstantial "Gothic principle," not a tangible physical trait. Nevertheless, Lassus's figurative St-Eustache is homologous with the more familiar perceptual structure/ornament building for which he is self-consciously offering a metaphorical evocation.

Furthermore, it would be a mistake to insist too much on Lassus's nonarchitectural imagery, for in the passage from which the above phrase is drawn he easily moves back and forth between figurative and literal architectural descriptive terms as he narrates what he perceives as the progressive decline experienced in French architecture since the introduction of classical forms (a story in which, therefore, the initial transitional period is without merit). This passage is worth citing at some length:

It is at the end of the fifteenth century, in that period when the art of the Middle Ages was in decline, that one begins to see the emergence of the antique element. In seeking to combine it with our national art, this element spawned all those monuments of a bastard style, such as Saint-Eustache, for example, where the Gothic principle exists in its entirety, but is completely denatured, weakened, and falsified by an envelope that is utterly alien to it. Up until that moment the antique element plays but a very secondary role. . . . But the closer one gets to our time, the more ambitious it becomes and it forges ahead until finally - master of the terrain, and having snuffed out even the slightest remains of Gothic art - it dares to show itself in all our monuments. . . . But what then is architecture? an assemblage of forms which are always in conflict with the construction; an incoherent whole where the least details, even though distorted, nevertheless betray their antique origin? It is there that one must search for the cause of the error in which the rationalist school has fallen. The followers of this school, struck as we are by the decadent state of our architecture, have understood, as we have, the necessity of a reform; but rather than searching for its rules in the genuinely true principles of our national art, they have not penetrated beyond the scaffolding of antique imitations which masked from them the very origin of those principles. Dashing off in pursuit of this false goal, which they have mistaken for the truth, even today they have not yet perceived that they have stumbled off the true path, that they have gone completely astray. The consequence is the error in which they have fallen as they thought that the serious study of ancient art would be sufficient to reform an architectural system, which is fundamentally corrupt, impossible in our climate, impossible with our materials. In dreaming of form, they have forgotten construction; in believing that they were creating art, they produced only archaeology.(81)

Lassus's opposition of essential, internal, true, and French construction to superficial, external, false, and foreign form articulates a fundamental assumption of much nineteenth-century architectural discourse: that a primary object of the gaze is to search for the contextually motivated essential principle of a building, which is not recuperable through a superficial scanning but through a penetration beyond mere external appearance. A building is theorized as having an inner, contextually responsive presence and an outer decipherable representation of that presence. In the best of cases the exterior form, contingent rather than essential, should exist in a motivated and expressive relationship to its internal essence: it is descriptive and makes manifest the inner presence, which is both structure and "structure." In bad (dishonest and immoral) architecture, such as is here described by Lassus, the exterior form obscures the interior principle either through excess or misrepresentation; it is a dangerous supplement, not good ornament.

This act of probing for a reality beneath the surface, of penetrating from the external to the internal, also informs the relationship posited between architecture and history. Just as ornament is understood as a sign of structure, as a contextually and materially deduced entity that should represent and point to a more genuine and prior inner principle, so too is architecture as a whole read as a product and sign of its times, of existing in an illustrative indexical relationship to its local and historically specific, material, topographic, and social context.

These progressive synchronic soundings are metaphorically both archaeological in their movement from surface to depth and architectural in their movement from exterior to interior. But I am describing more than just another example of the reliance of modern architectural discourse on historical (archaeology functioning as a historical tool) and architectural metaphors: each describes the condition of the sign that is metaphor in its opposition of an inner or buried real presence and an external surface representation of presence. In this figurative, symbolist conception of architecture, buildings are signifiers of the contexts that produced them and are themselves conceived as divided into representational signifiers, which are both ornament and "ornament," and essential signifieds, which are both structure and "structure."

Such a metaphorically resonant symbolic apprehension of architecture, with its imbrication of structure/ornament and "structure/ornament" and its mandate to penetrate to a true essence beyond the surface of a building, was not restricted to French theory. It was developed elsewhere as well, particularly in Germany (and later Vienna).(82) In the mid-1800s Carl Botticher and Gottfried Semper coined potent theoretical and metaphorical neologisms such as Bekleidung (dressing), Kernform (core or kernel-form), and Kunstform (art-form) to express the duality of the literally material and metaphorically extramaterial aspects of architecture that I have here outlined.(83)

In Botticher's Die Tektonik der Hellenen (1844-52) and "Das Prinzip der hellenischen und germanischen Bauweise" (1846), for example, Kernform simultaneously signifies the physical structure and the ontological essence of a building, which are responsive to material and function, while Kunstform signifies ornament as that contingent yet necessary part of a building that symbolically represents and makes culturally intelligible, through the traditional art function of mimesis, the underlying essence and functional structure of the building. The supplemental status of Kunstform is made clear when Botticher defines it is as the "explanatory layer. . . . [that] make [s] visible the concept of structure and space that in its purely structural state cannot be perceived"(84)

Through theorizing architecture in terms of its underlying ontological Kernform and its surface representational Kunstform, Botticher was able to narrate a history of architecture where Greek and Germanic Gothic emerged as the two high points, and where the German Renaissance was among those periods condemned for their use of dangerous dissembling ornament. Like Lassus writing about his national Renaissance, Botticher characterized sixteenth-century Germany as a period

when misunderstood antique forms were adopted to clothe buildings in the Germanic [Gothic] style. No lengthy critique of such a meaningless welter of forms is called for. . . . [This] school of thought . . . remained tied to the surface of things. No one realized that the origin of all specific styles rests on the effect of a new structural principle derived from the material and that this alone . . . brings forth a new world of art-forms.(85)

In much twentieth-century historical scholarship such theoretically sophisticated, metaphorically inflected (and parergonally entangled) understanding of historical architecture is often transformed into a direct, unmediated transcription of simple perceptual (if historically informed) knowledge, from which has been stripped all metaphorical signification and hence theoretical nuance. "Structure" and "ornament," Kernform, Kunstform, and Bekleidung have all become literal built structure and ornament.(86)

As it was originally articulated in the nineteenth century, structure/ornament resonates both as a literal description of architecture and as a theorized concept dependent on figurative meanings and interchangeable with figurative terms. Nineteenth-century writers are more fully conscious of what they are saying than are their later counterparts, who foreground structure/ornament as a strictly literal description of the material building (although in neither case is the spatialized, narrative function of the pair recognized). In other words, if twentieth-century historians simply write "Gothic structure/Renaissance ornament" when they look at St-Eustache with more predictable regularity than do their nineteenth-century predecessors, this shift in frequency of occurrence does not indicate a major epistemological shift: everyone is essentially seeing and saying the same thing, even if the earlier texts demonstrate a greater awareness of all that is being said as well as more of an alignment between authorial and textual meaning.

The insistent repetition of a single descriptive formula in the twentieth century, however, parallels a shift of another type: one in the primary location from which serious statements about the building are made. In the nineteenth century they tend to be found in architectural theory based on the study of history, whereas in the twentieth they appear in academic architectural history that has to a large extent forgotten its theoretical origins, which are not therefore a subject of critical inquiry.

Building "Structure/Ornament"

That structure/ornament has been naturalized as a common-sense way of describing much historical architecture has a great deal to do with the fact that it has also been naturalized as a way of building modern architecture. Lassus was not alone in advocating that a penetrative binary reading of historical architecture would benefit - even more, would save - the architecture of his time. The desire to search for contextually motivated essential "structures" recuperable under veils, clothing, shells, flesh, husks, shrouds, envelopes, and the like, as a means of revealing how nineteenth-century architecture should proceed was widely expressed by his contemporaries.(87) Vaudoyer, for example, before he turned. his attentions to the French Renaissance, wrote:

I thus think that in order to satisfy the needs of our time, one must preferably study the fundamental architecture of the ancients, that is to say, that which had to satisfy basic functions and was not yet corrupted by luxury. It is in this fundamental architecture that one can best rediscover the reason for forms, in a word, the skeleton, which later conceals itself under rich garments.(88)

As a consequence of such pedagogically motivated looking at historical architecture, contemporary architecture itself came to be conceived in terms of the structure/ornament opposition. Henri Labrouste's Bibliotheque Ste-Genevieve (1858-50), as Neil Levine has shown, was the first monument where this binary theory of contextually and functionally responsive architecture sought material reification.(89) This theory also informed the way the new architecture was looked at and talked about. Describing the method of a Labrouste follower a contemporary wrote:

He would turn his attention first to the skeleton and when he had weighed and balanced all its parts, he dressed his building as needs and function dictated and according to the resources at hand, but always allowing, under the folds of the attire, the means to divine a healthy and vigorous form.(90)

If structure/ornament would eventually come to be transformed into a literal description of historical architecture, it was almost immediately transposed to contemporary architectural production as architects tried literally to build it.

And, of course, twentieth-century architecture is also largely conceived according to this bipartite model. The emergence of structure as an independent and essential entity in the nineteenth century meant that it was able to assume a representational role that previously had been almost exclusively the domain of ornament. As a result, the potential superfluity of ornament and its slide into the merely, superficially, or decadently decorative was virtually inevitable, as were the endless debates about the relationship between structure and ornament, construction and form, tectonics and representation, technology and history in modern and modernist discourse.(91)

Whether a twentieth-century architect is for or against ornament, it continues to be understood as a separate, additional extra that is attached to the structure of a building (even if it is conceived as motivated by and representing an internal "structural" reality). This is true for architects who have taken the most polemical positions, such as those modernists who sought to strip buildings down to their aesthetically and ethically pure Kernform, or those postmodernists who wanted to build "decorated sheds" rather than architectural "ducks," and also for those who have sought to conceptually reconcile the two, such as Frank Lloyd Wright, with his Louis Sullivan-inspired "organic architecture," or those deconstructivist architects who regard the structure/ornament pair as a problematic oppositional modernist one that they are nevertheless obliged to work with or "between."(92) All such positions are heir to those nineteenth-century architects who made the initial cleavage as they sought to transform a theoretical construct, which opposed a building's essential principle to its surface garb, into built reality. Consequently, it is not only the multiple heuristic and figurative functions of the pair that sustain it as a persuasive structure in the St-Eustache literature but also its ubiquity in modern written and built architectural discourse. The structure/ornament St-Eustache is acceptable because we tend to think that's how buildings really work.

What is often forgotten, however, is that structure/ornament is a distinctly modern invention, as is the status assigned to its two members. Before the nineteenth century ornament was not paired with a tangible physical thing called structure; structure was not understood to be an entity with a self-sufficient ontological, representational, or aesthetic presence; and ornament was not reductively and exclusively conceived as a discrete, detachable object opposed to structure and subject to its own internal logic.(93) Rather, ornament had a more subtle, mobile, even incorporeal meaning.

Alberti, for example, in one of his many statements about ornament, insists, "The chief ornament in every object is that it should be free of all that is unseemly."(94) Unlike modern theorists who understand ornament to be the added presence of some material thing, Alberti here defines it as the absence of a quality, the unseemly. In De re aedificatoria ornament is a supplement not to structure but to beauty ("ornament may be defined as a form of auxiliary light and complement to beauty"(95)). Consequently, ornament is an extremely fluid term: whatever means can be brought to bear to reveal the inherent beauty of architecture all come under the heading of ornament. Something as tangible as a column or the leaves on a Corinthian capital are described by Alberti as ornament, but so too are the quality of the material and workmanship used for a building, its relative proportions, and the relationship between its component parts. Even empty spaces apertionum - in the surface of a wall are described as ornament.(96) Every component member of a building can both contribute to and constitute its formal beauty:

We must therefore take great care to ensure that even the minutest elements are so arranged in their level, alignment, number, shape, and appearance, that right matches left, top matches bottom, adjacent matches adjacent, and equal matches equal, and that they are an ornament [ornamentum] to that body of which they are to be a part.(97)

But if the "minutest element" can be ornament, it can also be part of structure (structura), of which Alberti writes, "It is not difficult to discover the parts that make up the structure: clearly they are the top and bottom, the right and left, the front and back, and all that lies in between."(98) That is, everything is ornament and everything is structure.

Looking at this issue from another perspective, for Alberti and many architectural theorists down through the end of the eighteenth century, a building is the result of two overlapping and necessary practices in which an architect must be equally skilled: construction and ornamentation. The first assures that a building will be well built, the second that it be beautiful and pleasing to the eye. But the two methods do not generate two classes of material traits: they do not produce structure/ornament.

In the context of the subject of this essay it is particularly relevant to cite Laugier's remarks on this methodological duality:

We understand decoration far better than they [Gothic architects] did: but they were far more skilled than we are in construction. If we want to improve ourselves, we should not consult them when the decoration of buildings is concerned, and we should never stop consulting them for the method by which they are constructed.(99)

This passage, from the Essai sur l'architecture, occurs in Laugier's discussion "On the stability of buildings," where he deplores the lack of practical and sophisticated construction knowledge on the part of contemporary architects and contrasts Gothic buildings, which have lasted for centuries, despite their being so light in construction, with recent works (such as St-Sulpice), which are unappealingly heavy and massively built yet are already displaying signs of structural weakness. Like Alberti, Laugier distinguishes between the formal shaping and proportioning of a building and the method of its construction. Thus, the construction methods employed by Gothic architects can be comprehended apart from the forms in which they are built, but no formal distinction between different classes of constructed versus decorated traits in any given building (that is, "structure" versus "ornament") is suggested.

An explicit statement on the essential and never entirely detachable character of that which is decorative appears in William Chambers's A Treatise on the Decorative Part of Civil Architecture. Like "father Laugier," as he calls his predecessor, Chambers adheres to the traditional distinction between. methods of construction and of decoration and writes of the latter:

The orders of Architecture . . . are the basis upon which the whole decorative part of the art is chiefly built, and towards which the attention of the artist must ever be directed, even where no orders are introduced. In them, originate most of the forms used in decoration; they regulate most of the proportions; and to their combination, multiplied, varied and arranged in a thousand different ways, architecture is indebted for its most splendid productions.(100)

The orders are so fundamental to the decorative part of architecture that they silently organize the forms of a building even when they are physically absent. Again, rather than a category of removable traits, the decorative signifies the entire formal repertoire of a building and the proportional relationships between its various elements.

If Laugier and Chambers speak of construction and decoration rather than construction and ornament, this does not mean that the word ornament has been abandoned. For these authors, "decoration" generally signifies what "ornament" did for Alberti, while "ornament" is now that which embellishes the orders (as was increasingly true in post-Albertian writing).(101) That is, "ornament" is applied to the orders (which, for Laugier, include "the Gothic order"), and the parergonal orders "not only decorate the building, but must constitute it."(102)

It is revealing to follow Laugier's difficulties as he attempts to define "ornament" in terms of detachability. He writes, "that which we call an ornament is an incidental adornment [parure] . . . which can be excised without harming that which is essential to the architectural order."(103) In the distinction Laugier then seeks to make between a column and a pilaster this definition is completely undermined. He initially argues that a pilaster cannot be ornament because it is "clearly an essential part of the architectural order, forming a whole with the entablature," and one cannot "remove it without corrupting the character of the composition." Yet he concludes his discussion on the pilaster by claiming that if pilasters and their entablatures were cut away from the masses against which they are "plastered," the building "would only lose its ornament." A pilaster order in ideal isolation may be "decoration" (and thus essential), but once it is actually applied to a building it is demoted to trivial, removable "ornament."

In a building that employs an order of freestanding columns, the contrary situation obtains: "one cannot touch a single one of its elements without damaging and ruining the building." But, according to Laugier, columns are also forms that should be applied in the place of pilasters: wherever a pilaster appears, he advises us, a column should be inserted in its place.(104) Consequently, a column is both "decoration" - a primary means of obtaining beauty in architecture - and has the character of "ornament," being something that can be added (and, by extension, taken away), and, following Laugier's logic, an element that cannot be essential to architectural beauty.

Laugier's nonrigorous and self-contradictory, confusing and confused attempt to distinguish column/decoration/essential from pilaster/ornament/incidental pointedly evidences that although prior to the nineteenth century ornament (or decoration) was never paired with and opposed to structure, the supplemental status of that which ornaments or decorates is a condition that tends to unite writings on the subject from the Renaissance (and earlier) to the present day. Any effort to distinguish ornament from what it ornaments will inevitably find itself on unstable ground, whether that which is ornamented is a physical structure, architectural beauty, or the orders as a compositional and formal ideal.(105)

Nevertheless, real epistemological differences can be recognized in the understanding of ornament and what it ornaments in texts before and after the early nineteenth century. For theorists such as Alberti, Laugier, and Chambers the concept of ornament contained in the following modernist pronouncement would have been utterly unintelligible:

[P]ositive quality or beauty in the International Style depends upon a technically perfect use of materials . . . upon the fineness of proportion in units such as doors and windows and in the relationship between these units and the whole design. The negative or obverse aspect of this principle is the elimination of any kind of ornament or artificial pattern. . . . Intrinsically there is no reason why ornament should not be used, but modern ornament, usually crass in design and machine-manufactured, would seem to mar rather than adorn the clean perfection of surface and proportion.(106)

The paring back of "ornament" or "decoration" (which are used interchangeably in the modern texts cited in this essay) to mean exclusively applied ornament, and the jettisoning from "ornament" of such traditional concerns as technical perfection, quality of materials, harmony and balance, and the "perfection of surface and proportion" was, of course, a crucial maneuver that allowed "structure" to materialize as a discrete entity. What is new to the modern period is not the slippery supplemental status of ornament but rather the effort to so narrow the definition of ornament that it becomes only (and problematically) detachable ornament. And ff ornament is to be detachable it must be detachable from something as concrete, autonomous, and meaningful as itself: the structure/ornament pair is thereby launched.

In failing to note the significant differences between processes of construction and decoration and structure/ornament and in unthinkingly accepting the structure/ornament pair as a ubiquitous architectural phenomenon, contemporary scholars have often unwittingly retroassimilated construction/decoration to the modern binary paradigm, as in this statement of 1988: "Alberti formulated for the first time the opposition between structure and ornament. . . . Certainly it has only been a devotion to the same distinction that has allowed architects in the last hundred years to deny the need for any ornament at all."(107) Thus, a premodern text is compelled to say something its author cannot even imagine, and the apparent elimination of ornament in twentieth-century architecture is wrongly understood as responding to a possibility that had been immanent for centuries, rather than since the early 1800s, when ornament first received its potentially detachable status.(108)

This does not mean that all scholars, particularly contemporary scholars of nineteenth-century architecture, are unaware of structure/ornament as emerging in a historically specific theoretical framework. Much has been written on this issue, and my above analysis makes use of this work. Yet although a great deal of this writing has been highly perceptive, in taking as one of its principal goals the undoing of modernist hostility to and misreadings of nineteenth-century architecture and theory, it often exhibits a marked tendency to duplicate uncritically the assumptions of its subjects. We are told, for instance,

Neo-Grec [Romantic] architects were the first to make the radical distinction between structural principle and decorative form. In demanding that forms of decoration be rationally induced from the materials and methods of construction as well as from the specifications of the program, the Neo-Grec architects acknowledged the distinction between appearance and reality as simply a matter of fact, and therefore saw the process of design as the decoration of construction. . . . The classical ideal of apparent formal homogeneity was replaced by the reality of structural differentiation. . . . In acknowledging a basic disjunction between form and content, the substantiality of structure and the insubstantiality of clothing form, the Neo-Grec offered a new literary syntax of expression. . . .(109)

In this presentation of architects who "acknowledge" and "realize" facts," in this concurring juxtaposition of an invalid and tendentious "classical ideal of apparent formal homogeneity" with "the reality of structural differentiation," and so on, we encounter a sympathetic, uncritical reader who is persuaded that his subjects have revealed eternal truths rather than manufactured conceptual structures that sustain a provisional epistemology.(110) Such interpretations serve to encourage the view that the structure/ornament opposition was a hard-won discovery of a self-evident and universally valid phenomenon, not the historically contingent invention it in fact is. So embedded in modern thinking is the pair that scholars of the French Renaissance (and of other periods traditionally seen as transitional, early, late, or peripheral) are not alone in continuing to reiterate it as they figure much historical architecture according to its rigid logic; they are joined by those who are more directly engaged with the strategies that produced this view or with the problematics it engenders.(111)

It may well be that architects cannot evade the authority of the modern structure/ornament construct (just as painters have never been able to escape completely another much earlier invention that is also often assumed to be a discovery, linear perspective).(112) But architectural historians are another matter, In this paper I have tried to suggest that in failing to recognize the historical origins, theoretical dimensions, and metaphorical and figurative guises of an apparently innocent description, contemporary historians, in dealing with buildings "like" St-Eustache, will find that the properly architectural continues to elude them as the form of buildings remains shackled to the strange burden of telling us their stories.(113)

Notes

For kindly reading my manuscript and their helpful comments on it I would like to thank Jean-Louis Cohen, Romy Golan, Linda Nochlin, Tilde Sankovitch, and Marvin Trachtenberg. Unless otherwise indicated, all translations are mine.

1. I should stress that this tends to be the case only in much contemporary architectural history, for the properly architectural is a concern of many contemporary architectural theorists. The periodical Assemblage, for example, includes a regular feature entitled "The Strictly Architectural," which "raises the question of what 'properly' belongs to architecture." For a summary and examples of writings on this issue, see also Kate Nesbitt, ed., Theorizing a New Agenda for Architecture: An Anthology of Architectural Theory, 1965-1995 (Princeton: Princeton Architectural Press, 1996).

2. Viollet-le-Duc, vol. 1, 240.

3. For instance, Jacques Du Breul wrote, "It will be one of the most beautiful buildings in Europe if it can be completed as it has been begun. For nothing is lacking as far as the perfection of the architecture is concerned, whether in regard to its great height, the windows and openings, or the enrichment of the diverse friezes and moldings, which are of all types and manners"; Du Breul, Le theatre des antiquitez de Paris (Paris: P. Chevalier, 1612), 793. And according to Germain Brice, "The building . . . is at present the largest and most spacious in the kingdom. The great size of the entire work, the number of piers (which are in truth a bit crowded), and the height of the vaults, with the chapels, which are all around; all these things together make this edifice magnificent"; Brice, Description nouvelle de ce qu 'il y a de plus remarquable dans la ville de Paris (Paris: La Veuve Audinet, 1684), vol. 1, 103. See also C. Le Maire, Paris ancien et nouveau (Paris: Theodore Girard, 1685), vol 1, 523; and Isaac de Bourges, Description des monuments de Paris (Paris, [16-]), vol. 1, Collection des anciennes descriptions de Paris, ed. Valentin Dufour (Paris: A. Quantin, 1878), 56. I would like to thank Paul Chenier of the Canadian Centre for Architecture for transcribing the Brice citation.

4. This rhetoric dominated architectural description until the advent of classical discourse, and St-Eustache was not the only building configured by it. It also informs, for example, the way Philibert Delorme looked at the architecture of ancient Rome, whose abundant diversity of architectural forms and motifs he enthusiastically described and illustrated. For example, he wrote about the Composite order, "on it was placed as much ornament and richness as possible, without leaving unadorned a single part of its cornice, cymas, astragals, echinus, crowns, dentils and all other members . . . including even the abacus of the capitals where were sculpted eggs and all sorts of friezes"; Delorme, Le premier tome de l'architecture (1567), repr. in Delorme, Traites d'architecture, ed. and commentary by Jean-Marie Perouse de Montclos (Paris: Leonce Laget, 1988), 201v. Similarly, if more succinctly, Arnold Van Buchel saw the Chateau of Madrid in the Bois de Boulogne as a superlative building that contained "a hundred rooms, with as many windows as there are days in the year"; Van Buchel, Description de Paris, 1585-86, in Memoires de la Societe de l'Histoire de Paris et de l'Ile-de-France 26 (1899): 87-88. And Sebastien Rouillard, looking at the porches of the cathedral of Chartres, wrote, "All of these portals are enriched with stories that are artfully delineated and sculpted. They are also embellished with an endless number of tall columns on which is placed such a quantity of statues of more-than-human size and of such admirable workmanship that they seem to disappear from sight because of the others - one scarcely knows at which one to pause or of which to have the highest opinion"; Rouillard, Parthenie, ou histoire de la tres auguste eglise de Chartres (1609), trans. in Robert Branner, ed., Chartres Cathedral (New York: W. W. Norton, 1969), 105.

5. This did, however, receive some methodological elaboration, particularly insofar as different phases of the Gothic were concerned; Jean Francois Felibien, for instance, sought to devise a rational, empirical methodology by which Gothique ancien (now labeled Romanesque or earlier) could be distinguished from Gothique moderne (Gothic proper), in his Recueil historique de la vie et des ouvrages des plus celebres architectes (Paris: La Veuve de Sebastien Marbre-Cramoisy, 1687). On Felibien's stylistic distinctions, see Middleton, 25: 299-301.

6. Such as Henri Sauval, Histoire et recherches des antiquites de la ville de Paris, written ca. 1660 (Paris: G. Moette, 1724), vol. 1, 437; and Jacques-Francois Blondel, Architecture francoise, vol. 1 (Paris: C. A. Jombert, 1752), 70.

7. Such as Michel de Fremin, Mires critiques d'architecture (Paris: Charles Saugrain, 1702), 27; and Jean Lebeuf, Histoire de la ville et de tout le diocese de Paris (1754), ed. and annotated by H. Gocheris, vol. 1 (Paris: A. Durand, 1863), 121-22.

8. Such as Germain Brice, Description nouvelle de ce qu'il y a de plus remarquable dans la ville de Paris, 3d rev. ed. (Paris: Nicolas Legras, 1698), vol. 1, 219; and Antoine-Nicolas Dezallier d'Argenville, Voyage pittoresque de Paris par M. D(Paris: Les freres de Bure, 1778), 164.

9. Marc-Antoine Laugier, Observations sur l'architecture (Paris: Desaint, 1765), 150.

10. Ibid., 283-84. The comparison Laugier is here making is similar to that which he had made earlier between the good Gothic Notre-Dame and the bad modern St-Sulpice; Laugier, 174-75.

11. Indeed, such attempts are largely inarticulate. Laugier's one effort to see such duality is notable in that he managed to perceive at least one morphologically identifiable feature (classical colonnettes) at St-Eustache. More typically, we are simply informed that the building is Gothic and antique and are not told what specific elements at St-Eustache adhere to either manner. Antoine-Nicolas Dezallier d'Axgenville (as in n. 8), for example, writes "Its architecture is of a Gothic sort mixed with a bad antique." More expansive, yet equally elusive, is Germain Brice's figuration of the building in the 1698 third edition of Description nouvelle, which differs markedly from that offered in the first edition of 1684 (see n. 3 above). Whereas he had earlier praised St-Eustache as a superlative example of Aristotelian magnificence (a critical judgment supported by such specified traits as the quantity of chapels and the height of the vaults), Brice (as in n. 8) now writes: "The architect has made a horrible confusion of Gothic and antique, and has so corrupted one and the other that nothing regular or tolerable can be seen, so that one must regret, with justice, the great sums of money that were spent on the building under the direction of the miserable mason who gave the designs." Unlike virtually every other building in Paris, which Brice is able to describe with varying degrees of closely observed detail, St-Eustache has become an elusive structure, intractably refusing to yield to the cognitive and descriptive gestures of the classical gaze. Language, the necessary and transparent analytical tool of classical discourse, which should automatically know and be able to name what it sees, is here rendered impotent. Furthermore, I would argue, it is this new, frustrating (and singular) strangeness that causes Brice's condemnation as he executes a metonymic shift from method to object and displaces his disorientation and confusion onto the building: "I am confused, therefore the building must be confusing."

12. Whereas 19th- and 20th-century observers of St-Eustache could look, for instance, at its piers and see elements of a Gothic structure covered with Renaissance ornament, in the early 1700s, Michel de Fremin (as in n. 7), who paired St-Eustache and St-Sulpice together as examples of bad modern architecture, saw the piers of these buildings as morphologically unified and contrasted their quantity and massiveness to the lighter columnar Gothic supports of Notre-Dame (27-39). For Fremin a pier was Gothic or not, classical or not: it could never be both. For modern descriptions of the piers, see [A.] Le Roux de Lincy (text) and Victor Calliat (plates), Eglise Saint-Eustache a Paris (Paris: By the authors, 1850), 20; and Earl Rosenthal, "The Diffusion of the Italian Renaissance Style in Western European Art," Sixteenth Century Journal 9 (1978): 36.

13. Thus, neither the 18th-century architecture (and the theories in support of it) that Robin Middleton has dubbed "Graeco-Gothic," where Gothic methods of construction are entirely realized in classical forms (as at Ste-Genevieve), nor those 18th-century projects to "improve" Gothic architecture through the reshaping of their forms as classical are able to foster or accommodate a binary structure/ornament recognition of St-Eustache. On 18th-century theories of Gothic architecture, see Middleton; and Wolfgang Herrmann, Laugier and Eighteenth Century French Theory (London: A. Zwemmer, 1962; reprint, 1985), chaps. 5-7. The formal rectification of the Gothic is discussed by Laugier (as in n. 9), 129-51. For more on the distinction between 18th-century construction and 19th-century structure, see the section below entitled "Building 'Structure/Ornament.'"

14. Vachon, 149-50; Rene Schneider, L'art francais: Moyen age a Renaissance (Paris: H. Laurens, 1923), 164-65; W. H. Ward, The Architecture of the Renaissance in France, 2d rev. ed. (London, 1926; reprint, New York: Hacker Art Books, 1976), vol. 1, 85; Pierre Lavedan, L'architecture francaise (Paris: Larousse, 1944), 103; Yvan Christ, Eglises parisiennes, actuelles et disparues (Paris: Editions "Tel," 1947), 43; Blunt, 33; Amedee Boinet, Les eglises parisiennes, vol. 1 (Paris: Editions de Minuit, 1958), 471; Rosenthal (as in n. 12), 36; Michael Hesse, Von der Nachgotik zur Neugotik: Die Auseinandersetzung mit der Gotik in der franzosischen Sakralarchitektur des 16ten, 17ten, und 18ten Jahrhunderts (Frankfurt: Peter Lang, 1984), 25; Henri Zerner, "Le frontispiece de Rodez: Essal d'interpretation," in "Il se rendit en Italie": Etudes offertes a Andre Chastel (Paris: Flammation; Rome: Edizioni dell'Elefante, 1987), 303; Jean-Marie Perouse de Montelos, Histoire de l'architecture francaise de la Renaissance a la Revolution (Paris: Editions Menges, 1989), 73-74; Willibald Sauerlander, review of L'art de la Renaissance en France: L'invention du classicisme, by Henri Zerner, New York Review of Books, Oct. 9, 1997, 47.

15. Dieter Kimpel, Paris, Fuhrer durch die Stadtbaugeschichte (Munich: Hirmer, 1982), 160; Hesse (as in n. 14), 25; David Thomson, Renaissance Paris: Architecture and Growth 1475-1600 (Berkeley: University of California Press, 1984), 187; and Perouse de Montclos (as in n. 14), 74-75.

16. Spiro Kostof, A History of Architecture (New York: Oxford University Press, 1985), 403; and David Watkin, A History of Western Architecture (New York: Thames and Hudson, 1986), 228. On the non-Italian Renaissance in general, Kostof writes, "most commonly budding types current in the host country would wear trinkets, or even whole mantles in the new taste" (430); and according to Watkin, "outside Italy, Gothic persisted throughout the fifteenth century, knowledge of the Renaissance arriving after 1500 largely in the form of ornamental details" (210). The same view is presented in Peter Murray's survey, where we are told that in England, "as elsewhere in Europe, Italian influence came in by means of decorative details, and it was many years before the structural principles were understood, let alone copied"; Murray, Renaissance Architecture (New York: Rizzoli, 1985), 180.

17. Ludwig H. Heydenreich and Wolfgang Lotz, Architecture in Italy 1400-1600 (Harmondsworth, Eng.: Penguin, 1974), 100.

18. See, however, n. 111 below.

19. Christopher Wilson, The Gothic Cathedral (London: Thames and Hudson, 1990), 117.

20. Henri Focillon, The Art of the West, vol. 2, Gothic, trans. D. King, ed. Jean Bony (Ithaca, N.Y.: Cornell University Press, 1980), 148. It should alan be noted that Focillon observes the same obscuring process occurring in writings about art: "the luxuriant vegetation with which interpreters decorate the work of art accumulates around it, sometimes to the point of entirely concealing it from us." Focillon, Vie des formes (1934; 6th ed., Paris: Presses Universitaires de France, 1996), 2. Thus, it can be argued that Focillon understands Flamboyant architecture to be both a critique as well as a perversion of good 13th-century Gothic.

21. Vachon, 11.

22. Ibid., 149-50.

23. Vachon also avoids the macabre imagery of Viollet-le-Duc by substituting for squelette (skeleton) the more benign ossature, a word that also translates into the English "skeleton," and indeed has the same anatomical meaning and shares its figurative meanings of "framework" or "structure" but does not possess squelette's connotations of unwholesome creepiness, decadence, and morbidity. Viollet-le-Due certainly plays to these connotations here and elsewhere in the Dictionnaire in describing the appearance of "foreign," usually Roman, forms in French architecture, as when he writes of the early Middle Ages (vol. 1, 122), "architecture remains enveloped in its musty antique shroud."

24. According to Vachon (7), "There is a direct, immediate and uninterrupted lineage from the architecture of the Middle Ages to that of the Renaissance. The one comes from the other just as a plant grows from fruit sown in the ground. When it is a germ it is nourished by it; then, having established its roots, it draws fertilizing elements from the soil, sprouts, comes forth, grows; and finally, in the air and light - sources of fecundity, of life, and of beauty - it blossoms and bears fruit."

25. On the interpretive function of plots - distinguished from stories see Hayden White, Tropics of Discourse: Essays in Cultural Criticism (Baltimore: Johns Hopkins University Press, 1985), 58-61.

26. Blunt, 33.

27. It is worth noting that Blunt uses the term Renaissance to describe the architecture of 16th-century France only once in his entire discussion of the period (9, for the staircase at Blois). More typically, 16th-century buildings are characterized as less or more classical; that is, the period is largely seen as a nondistinctive one, meaningful only insofar as it anticipates French classical architecture of the 17th century.

28. The "superannuated" interpretation of 16th-century French church architecture was particularly popular with English historians before Blunt. For instance, in his chapter entitled "Church-Building in the Sixteenth Century, and the End of Gothic Architecture," Reginald Blomfield speaks of the "last flicker of medievalism," "the prolonged struggle between the old and the new," the "tendency to slip back to the old manner," of a "throwback to the motives of much earlier work," and so on; Blomfield, A History of French Architecture from the Reign of Charles VIII till the Death of Mazarin (London: G. Bell and Sons, 1911),vol. 2, 1, 6, 11.

29. Eugene-Emmanuel Viollet-le-Due, "Saint-Eustache," in Para Guide, par les principaux ecrivains et artistes de la France (Paris: Librairie Internationale, 1867), vol. 2, 692.

30. Sauerlander, 4.

31. Viollet-le-Due, for example, wrote about the secular architecture of 16th-century France (vol. 1, 327): "The great French architects of the sixteenth century . . . were able to unite with remarkable skill the tried and true traditions of past centuries with the recently accepted forms. If they employed the antique orders, and if they often believed themselves to be imitating Roman art, nevertheless in their buildings they respected the needs of their time and submitted themselves to the requirements of climate and materials." See also idem, Entretiens sur l'architecture, vol. 1 (Paris: A. Morel, 1863), 338-39. Unlike Viollet-le-Duc's interpretation, in many 19th-century readings of the French Renaissance the emphasis on local requirements often served the anxious nationalistic agenda of minimizing or denying the presence of Italian influence. Thus, for example, about the Paris Hotel de Ville (designed by Domenico da Cortona), Leonce Reynaud wrote, "Remarkably, nothing evokes Italy in this architecture designed by an Italian artist. . . . the decorative system presents that charming fantasy, that freedom of manner, which properly belongs to the French Renaissance and never passed to the other side of the Alps"; Reynaud, Traite d'architecture contenant des notions generales sur les principes de la construction et sur l'histoire de l'art, vol. 2 (Paris: V. Dalmont, 1858), 421. This anti-Italian interpretation of French Renaissance architecture received its most extreme and poignant form in Leon Palustre's La Renaissance en Franee, vol. 1 (Paris: A. Quantin, 1879), ii-iii, where it is claimed that the very term Renaissance is inappropriate when applied to France because "Unlike Italy, France did not slumber after the fall of the Roman Empire; it was not necessary therefore to wake her up but simply to set her on a new path. The rupture that is assumed to separate the present and the past never existed, and nothing that was created on this side of the Alps, particularly during the first half of the sixteenth century, is in absolute opposition to the manner that had long been in practice. . . . The so-called Italian domination of our sixteenth-century architecture makes as much sense as the influence attributed to the Goths during the most beautiful period of the Middle Ages."

32. Sauerlander, 5.

33. Ibid., 11.

34. This is not to suggest that the problematics of the transitional can be escaped by simply sidestepping or deconstructing the structure/ornament concept. To unravel "transitional," however, is beyond the scope of this paper. Here I will only observe that the primary theoretical problem posed by "transitional" is that, even within the internal logic of this mode of conceiving architecture, the supposed "opposite" of a transitional building, a monument that conforms to a norm, does not in fact exist. Normative models can never be iconically represented in any real building, for they are always absent, always different from the real building; they are origins always lost in the past or ends always deferred to the future. It is only because movement is conceived as present in each individual building, which transcribes its absent origin and deferred end, that the modern observer can simultaneously think of an individual building and the evolution of architecture of which it is a part. Hubert Damisch has observed about Viollet-le-Duc: "the 'ideal,"complete,"achieved,' model which he proposes for the Gothic cathedral, no architect ever had it in mind at any moment of history; and, if it appears in the Dictionnaire, it is precisely as a model, a conceptual tool well suited to reveal the structural ties that unite the multiple creations of the Gothic age, the shared field where they successively appear, and the problematics of the whole that was at the horizon of each singular experience." H. Damisch, introduction to L'architecture misonnee: Extraits du Dictionnaire de l'architecture francalse by Eugene-Emmanuel Viollet-le-Due (Paris: Hermann, 1964), 19. When the transitional building draws attention to the apparently greater expanse of its conceptual horizon it is illuminating a panoramic prospect equally available to and equally present in every building. In the evolution of the history of architecture as conceived by modern historians buildings can only be positioned spatially and temporally closer to an ideal model and further away from another: the point that marks a break in the continuum between a normative and transitional period is always an arbitrary one. "Transitional," in other words, is the normative condition of any building located in this formalist continuum. On presence as the effect of spatial and temporal differences, see Jacques Derrida, "Differance," in Margins of Philosophy, trans. A. Bass (Chicago: University of Chicago Press, 1982), 3-27.

35. See, however, Denis Hollier, "The Architectural Metaphor," in Against Architecture: The Writings of Georges Batallie, trans. B. Wing (Cambridge, Mass.: MIT Press, 1989), 14-56.

36. Michel Foucault, The Order of Things: An Archeology of the Human Sciences, trans. A. Sheridan (New York: Vintage Books, 1973), 219.

37. Most recent scholarship on 19th-century architecture and theory takes up the theme of the new importance of history, historicism, and historicized interpretation; a particularly useful analysis of the historical turn in France is offered by Robin Middleton, "The Rationalist Interpretations of Classicism of Leonce Reynaud and Viollet-le-Duc," AA Files 11 (1986): 29-48. The significance of history to 19th-century architectural production is made clear by the architect of the Paris Opera, Charles Garnier: "architects who build monuments must consider themselves to be the writers of future history; they must indicate in their works the characteristics of the time in which they create; finally they must, through duty and through the love of truth, inscribe in their buildings those indisputable signs of the period of construction"; Garnier, "La reconstruction des monuments de Paris," Le Temps, Sept. 7, 1871, quoted in Christopher Curtis Mead, Charles Garnier's Paris Opera: Architectural Empathy and the Renaissance of French Classicism (New York: Architectural History Foundation, 1991; distributed by MIT Press, Cambridge, Mass.), 7 (my translation). Thus, viewers of architecture are not alone in searching for the signs that reveal the historical location of a building; contemporary buildings themselves must actually be constructed in such a way that their historicity in all its implications - as a sign of the present that will be legible in the future, as distinct from its past and future - is carefully encoded.

38. Barry George Bergdoll, "Historical Reasoning and Architectural Politics: Leon Vaudoyer and the Development of French Historicist Architecture," Ph.D. diss., Columbia University, 1986, xvi.

39. As Jacques Derrida has observed, "Now, stricto sensu, the notion of structure refers only to space, geometric or morphological space, the order of forms and sites. Structure is first the structure of an organic or artificial work ..]. the architecture that is built and made visible in a location. . . . Only metaphorically was this topographical literality displaced in the direction of its Aristotelian and topical signification"; Derrida, Writing and Difference, trans. A. Bass (Chicago: Chicago University Press, 1978), 15-16.

40. Ibid., 17.

41. Although Vaudoyer is sometimes cited as the sole author of the "Etudes" (Van Zanten, 1987, 325) and is given top billing by others (Bergdoll, 116-17), I am here adhering to the way the authorship is actually designated: "MM. Albert Lenoir et Leon Vaudoyer"; Lenoir and Vaudoyer, 7 (1839): 4. Vaudoyer published under his own name a much abbreviated version of the "Etudes," which repeated phrases, whole passages, and images of the joint endeavor but which continued through to his own day, whereas the "Etudes" concluded with the reign of Louis XIV; Vaudoyer, "Histoire de l'architecture en France," in Patria: La France ancienne et moderne, ed. J. Aicard et al. (Paris: J.-J. Dubochet, Lechevalier, 1847), vol. 2, 2113-09.

42. Lenoir and Vaudoyer, 8 (1840): 62-63.

43. Ibid., 63.

44. Lenoir and Vaudoyer, 12 (1844): 259-62; 14 (1846): 105-6.

45. Lenoir and Vaudoyer, 12 (1844): 260.

46. Ibid.

47. Indeed, their account of the 16th century stands in marked contrast to their history of earlier church architecture, of which they wrote in conclusion, "It is thus that we have seen . . . the Christian church, modest and simple at its birth, grow and acquire a noble severity in the eleventh and twelfth centuries, rise to the greatest height of its splendor in the two following centuries, begin to decline in the fifteenth, and finish by dying out in the sixteenth century"; Lenoir and Vaudoyer, 8 (1840): 63.

48. As David Van Zanten has written about Vaudoyer and his contemporaries, "In architecture, the basis of theorizing was not a temporally static interpretation . . . but was instead a historical interpretation of transitions - of syntheses - meant to define the mechanism of evolution in building"; Van Zanten, 1977, 223; see also 223-30; Middleton (as in n. 37); and Bergdoll, 120-25. According to Willibald Sauerlander (2), "Up to the end of the eighteenth century transition was a terminus technicus for certain forms of [literary or musical] composition. . . . It is only with the rise of modern historicism that transition . . . changed . . . into a term of evolution."

49. Already in 1837 Vaudoyer had expressed the opinion that "the true national architecture is that which was given pride of place during the reigns of Francis I and Henry II . . . this is the only one that can constitute modern French art, for . . . this architecture is much better suited to the needs of our present civilization than Gothic architecture whose forms . . . contrast notably with our customs and purposes"; Vaudoyer, ms., quoted in Bergdoll, 139 (my translation). On Vaudoyer's interest in the architecture of the French Renaissance, see Bergdoll, 131-40, 154-56, and passim. In the "Etudes d'architecture," the beginning of the end of the Renaissance was signaled in such works as Jean Bullant's Chateau of Ecouen, of whose colossal order Lentir and Vaudoyer wrote, "Here is where the Renaissance is truly vulnerable to attack; it is when its blind love of antiquity, which paralyzes all invention, lures it beyond the point of reason to produce those puerile imitations, the consequences of which became so quickly fatal"; Lenoir and Vaudoyer, 11 (1843): 298. Lenoir and Vaudoyer were far from alone in their interest in and advocacy of French Renaissance architecture as a model for contemporary building. From about the mid-century onward, for example, numerous folio-size, multivolume, lavishly illustrated studies of French Renaissance architecture appeared, such as Adolphe Berty, La Renaissance monumentale en France: Specimens de composition et d"ornementation architectoniques empruntes aux edifices construits depuis le regne de Charles VIII jusqu'a celui de Louis XIV, 2 vols. (Paris: A. Morel, 1864); Alfred Darcel and Eugene Rouyer, L'art architectural en France depuis Francois ler jusqu'a Louis XIV: Motifs de decoration interieure et exterieure dessines d'apres des modeles executes et inedits des principales epoques de la Renaissance, 2 vols. (Paris: Noblet et Baudry, 1863-66); Cesar Daly, Motifs historques d'architecture et de sculpture d'ornement pour la composition et la decoration exterieure des edifices publics et travis: Choix de fragments empruntes a des monuments francais du commencement de la Renaissance a la fin de Louis XVI, 2 vols. (Paris: Ducher, 1870). As their titles suggest, these works had a dual purpose, which Daly explicitly spells out at the beginning of his text (vol. 1, 1): "that of facilitating the practical and day-to-day work of the architect, and that of throwing new light upon the history of French architecture from the beginning of the Renaissance up to the present day." For a discussion of architectural periodicals and the French Renaissance, including an analysis of the modes of graphic representation utilized, see Francoise Beudon, "Le regard du XIXe siecle sur le XVIe siecle francaise: Ce qu'ont vo les revues d'architecture," Revue de l'art 89 (1990): 39-56.

50. Lenoir and Vaudoyer, 10 (1842): 126-27. On the use of the fragments of Gaillon in Felix Duban's courtyard at the Ecole des Beaux-Arts (1832-40); see Van Zanten, 1987, 71-83.

51. Thomson (as in n. 15), 188.

52. Leonce Reynaud, "Architecture," in Encyclopedie nouvelle, ed. P. Leroux and J. Reynaud, 2d ed. (Paris, 1839), vol. 1, 777; trans. in Van Zanten, 1977, 228. For Vaudoyer and Lenoir's take on the early Italian Renaissance, see Lentir and Vaudoyer, 10 (1842): 123-25.

53. In his Traite d'architecture Reynaud (as in n. 31) does not even try to cope with the issue when he charts the history of church architecture from the early Christian style latin through the eighteenth-century style moderne (218-317). The discussion of the style de la renaissance (301-11) contains no mention of French architecture, and St-Eustache appears retrospectively in the discussion of style mode, where it stands in for all church building of the otherwise ignored French Renaissance: "After a few attempts - of which the church of Saint-Eustache in Paris is the most complete and best realized example - to apply the forms of Renaissance architecture to the plans and proportions established by Gothic architecture, the basilica of Saint-Peter's became a model for the entire Christian world" (311). This sentence represents the sum total of Reynaud's thoughts on the history of ecclesiastical construction in 16th-century France, despite his earlier announcement that the French Renaissance will be one of the principal architectural styles he will consider (92), which he indeed otherwise does, as the secular architecture of the period presents him with no theoretical or narrative problems.

54. Lenoir and Vaudoyer, 10 (1842): 125-28, 193-200, 225-27, 265-68; 11 (1843): 49-54, 121-26, 193-98, 297-302, 397-401.

55. Lenoir and Vaudoyer, 11 (1843): 194.

56. Abbe Balthasar, "L'eglise Saint-Eustache de Paris," Revue Archeologique 11 (1855): 718. Similarly, according to Abbe Gaudreau, "the architecture of Saint-Eustache is precisely that of the transition of the Gothic to the Renaissance, where one finds the sober simplicity of the first style and the elegance of the second, which subsequently became overburdened with a reprehensible profusion of decoration"; Gaudreau, Notice descriptive et historique sur l'eglise et la paroisse Saint-Eustache de Paris (Paris: Dentu, 1855), pt. 2, 8.

57. Lenoir and Vaudoyer, 12 (1844): 259-60; 14 (1846): 106. As for the contention that the Italians had a natural affinity for the antique that led them to embrace fully its revived forms, the rather obvious counterclaim, which was then being championed by Viollet-le-Duc and others - that the French had a natural affinity for the Gothic that should have compelled them to reject the antique - was one that Lenoir and Vaudoyer were specifically arguing against. They take the position that 13th-century Gothic was a decadent corruption of good Christian Romanesque architecture and not a distinctively French mode of architecture worth reanimating. They thus dramatically denounced the ambitions of Gothic revivalists: "What! the Gothic should be our national art! and we should repudiate all that has been achieved since then! What! such would be the limits imposed on the French genius, and since the fifteenth century our art should have lost ali originality, all character! We cannot believe it"; Lenoir and Vaudoyer, 12 (1844): 262.

58. Lenoir and Vaudoyer, 14 (1846): 106. The text continues, "It is thus entirely natural that it was first in residential buildings that were adopted those alterations whose goal was to obtain a well-being and pleasures more in harmony with the civilization of that period." It was also for this reason that civic architecture - although not problematically atransitional - did not experience the building boom that residential architecture did; Lenoir and Vaudoyer, 10 (1842): 195.

59. Lenoir and Vaudoyer, 10 (1842): 121.

60. Blunt, 33.

61. Blunt is perhaps referring to such factors as those evoked by W. H. Ward (as in n. 14), 85: "whereas in the castle the reason of its being - its fortification - was growing obsolete, and some semblance of it was retained only from habit, in the church the functions were unaltered, and no change in essentials was tolerated."

62. In the preface to the fourth edition of Art and Architecture in France (1980) Blunt defined what he meant by classical: "In architecture I take the word 'classical' to imply the correct use of the Orders according to the practice of the Ancients . . . but at the same time the pursuit of certain qualities of clarity and simplicity, a preference for regular forms (circle and square), for plane surfaces, clearly defined masses, and simple materials such as stone and stucco rather than marbles and gilding, the result being a static monumental style related to certain familiar types of ancient Roman buildings . . . and to the works of Italian architects of the High Renaissance" (12).

63. Blunt, 32-34.

64. After a traditional structure/ornament critique of the building, A. Le Roux de Lincy (as in n. 12), for example, writes, "In spite of everything, it must be admitted that this monument has grandeur; its structure is strong and bold; the choir in particular has a quality that is austere and dignified" (20). Similarly, Leon Palustre (as in n. 31) writes, "If the ornamentation is sometimes capricious, the church never fails to be full of nobility and grandeur, so that certain features, which a strict logic cannot but condemn, effect an intense seduction upon the spirit" (vol. 2, 1881, 130).

65. The Lewis and Short Latin Dictionary, 1st ed., gives as one definition of structura "a building, erection, edifice, structure."

66. American Heritage Dictionary (1971); Webster's New International Dictionary, 2d ed.; Oxford English Dictionary, 1st ed.; emphasis added. These meanings are also signified by the Latin ornamentum, which the Lewis and Short Latin Dictionary, 1st ed., defines as "apparatus, accoutrement, equipment, furniture, trappings, etc. . . . an ornamental equipment, ornament, mark of honor, decoration, embellishment,jewel, trinket."

67. Leon Battista Alberti, On the Art of Building in Ten Books, trans. J. Rykwert, N. Leach, and R. Tavernor (Cambridge, Mass.: MIT Press, 1988), 261-62.

68. For some French examples, see n. 49 above.

69. On the problematics of the supplement, see Jacques Derrida, Of Grammmatology, trans. G. H. Spivak (Baltimore: Johns Hopkins University Press, 1974).

70. Jacques Derrida, The Truth in Painting trans. G. Bennington and I. McLeod (Chicago: Chicago University Press, 1987), 59-60.

71. Paul Frankl, Gothic Architecture (Baltimore: Penguin, 1962), 213-14.

72. Hesse (as in n. 14), 25; emphasis added.

73. Nikolaus Persner, John Fleming, and Hugh Honour, A Dictionary of Architecture (Woodstock, N.Y.: Overlook, 1976), 212.

74. On the role of architecture and architectural figures in philosophical and other discourses, see Derrida (as in n. 39), 3-30, 278-93; and Mark Wigley, The Architecture of Deconstruction: Derrida's Haunt (Cambridge, Mass.: MIT Press, 1993).

75. Hollier (as in n. 35), 33-35.

76. "The Pleasures of the (Hyper)text," Talk of the Town, New Yorker, June 27-July 4, 1994, 44.

77. Eugene-Emmanuel Viollet-le-Duc, "Du style gothique au XIXe siecle," Annales Archeologiques 4 (1846): 350-51.

78. Ibid., 341-43.

79. Derrida (as in n. 70), 64; Derrida (as in n. 69), 215.

80. On the antiphenomenological nature of Viollet-le-Duc's theory in the Dictionnaire, see Damisch (as in n. 34), 20-23.

81. Jean-Baptiste Lassus, "De l'art et de l'archeologie," Annales Archeologiques 2 (1845): 198-99.

82. For a critical overview of the international scene, see Kenneth Frampton, Studies in Tectonic Culture: The Poetics of Construction in Nineteenth and Twentieth Century Architecture (Cambridge, Mass.: MIT Press, 1995), chaps. 2, 3. In England, A. Welby Pugin's theories of Gothic architecture, where each ornamental motif was seen to assume or display a structural function, led him to an assessment of St-Eustache and the architecture of 16th-century France that strikingly parallels many contemporary French texts: "from the moment the Christians adopted this fatal mistake, of reviving classic design, the principles of architecture have been plunged into miserable confusion. . . . At first it was confined to the substitution of a bastard sort of Italian detail to the ancient masses. This is particularly striking in the French buildings erected during the reign of Francis the First. . . . The church of St. Eustache, at Paris, is a most remarkable example of this period"; Pugin, An Apology for the Revival of Christian Architecture in England (London: J. Weale, 1843), 7-8.

83. On Botticher (including a discussion of his tectonic theory in the light of contemporary debates in German aesthetic philosophy about the representational in art, the purposiveness of art, and the hierarchies of the arts), see Mitchell Schwarzer, "Ontology and Representation in Karl Botticher's Theory of Tectonics," Journal of the Society of Architectural Historians 52 (1993): 267-80. On Semper's theory of Bekleidung, see Harry Francis Mallgrave, Gottfried Semper: Architect of the Nineteenth Century (New Haven: Yale University Press, 1996), 180-81, 185-88, 293-302. On the relationship between the theories of Botticher and Semper, see Mallgrave, 219-25; and Wolfgang Herrmann, Gottfried Semper: In Search of Architecture (Cambridge, Mass.: MIT Press, 1984), pt. 2, chap. 3. For Semper, see also Gottfried Semper, The Four Elements of Architecture and Other Writings, trans. H. F. Mallgrave and W. Herrmann, intro. W. Herrmann (Cambridge: Cambridge University Press, 1989). The theories and neologisms of Botticher and Semper were widely taken up and developed by theorists and architects down to the end of the century, such as, for example, Adolf Loos, "The Principle of Dressing" (1898), in Adolf Loos, Spoken into the Void: Collected Essays 1897-1900, trans. J. Newman and J. Smith (Cambridge, Mass.: MIT Press, 1982), 66-69. For the history and application of one metaphorical neologism, see Werner Oechslin, "The Evolutionary Way to Modern Architecture: The Paradigm of Stilhulse und Kern," in Otto Wagner: Reflections on the Raiment of Modernity, ed. Harry Francis Mallgrave (Santa Monica, Calif.: Getty Center, 1993), 363-410.

84. Carl Botticher, "The Principles of the Hellenic and Germanic Ways of Building with Regard to Their Application to Our Present Way of Building," in W. Herrmann, trans. and ed., In What Style Should We Build? (Santa Monica, Calif.: Getty Center, 1992), 163; emphasis added.

85. Ibid., 153. Unlike Viollet-le-Due and Lassus, however, who in their Annales Archeologiques essays advocated a return to a Gothic structure purged of its classical scaffolding, Botticher proposed a different solution for contemporary architecture: "The structural principle is ... to be adopted from the [Germanic] arcuated system and transformed into a new and hitherto unknown system [through the exploitation of the new material of iron]; for the art forms of the new system, on the other hand, the formative principles of the Hellenic style must be adopted in order to give artistic expression to the structural forces within the parts, their correlation, and the spatial concept" (159).

86. An interesting reversal of this process occurs in Oechslin (as in n. 83), where, as indicated in the title ("The Evolutionary Way to Modern Architecture: The Paradigm of Stilhulse und Kern"), a 19th-century metaphor for structure/ornament is revived by the author and employed to chart a history of early modernism. Oechslin describes this history as novel in that it does not emphasize radical breaks with the past but rather stresses an evolutionary development from historicism (with its historicized ornament or Stilhulse) to pure modernism (with its naked structural Kern). In other words, Oechslin revives Stilhulse und Kern as a metaphor that allows him to narrate the story of early modernism, a story in which Otto Wagner becomes a transitional figure as the chronology of his oeuvre synecdochically embodies the movement from historicism to modernism. That the model is not only binary and sequential but also metaphorically organic (unlike the architectural structure/ornament) only strengthens its narrative capability. The result is a curious elision of 19th-century writing about itself and Oechslin's own narrative. He thus cites, for example, as remarkably prescient and "visionary" an 1886 text of Joseph Bayer, who described the path contemporary architecture will probably take: "now the mysterious vital forces push up and the real, true, and essential building form of the period grows powerful limbs within the traditional masks and draperies of style. And if in the end it is completely organized and fully mature, the so beautifully ornamented, historical Stilhulsen will peel away; they are shed forever and the new Kern appears bright and clear in the sunlight" (385).

87. On this new mode of interpreting architecture in France, see Levine, 357-93; and Van Zanten, 1987, chap. 2.

88. Leon Vaudoyer, letter of 1831; trans. (with slight modifications) in Van Zanten, 1987, 8. Similarly, Botticher (as in n. 84) wrote, "All opinions for or against a particular style have referred only to the outer shell, that is to the scheme of the building's art-forms, which were considered to be identical with its principle of style. The true essentials have never been seriously considered; the discussion has never actually turned to the source of the art-forms and of the diversity of styles, namely the structural principles and material conditions on which each is based" (150).

89. Levine, 325-57.

90. L. Radoux, in Revue generale de l'architecture et des travaux publics (Paris, 1879), 80; trans. in Van Zanten, 1977, 214. Criticism of contemporary architecture was also often framed in these binary material terms, as when Viollet-le-Duc (vol. 8, 493-94), wrote: "For many people, style in architecture only consists in a decorative envelope, and, even among artists, there are many who sincerely believe that they are producing a work of style because they have stuck several Etruscan, or Greek, or Gothic, or Italian Renaissance moldings or ornaments onto a structure that has nothing in common with the art of those periods."

91. On the mutation of ornament into the decorative, see Jacques Soulillou, Le decoratif (Paris: Klincksieck, 1990), 15-35.

92. The criticism of these positions often sounds a familiar note, as in, for example, this denunciation of postmodern "decorated sheds": "not only is there a total schism between the inner substance and the outer form, but the form itself either repudiates its constructional origins or dissipates its palpability"; Kenneth Frampton, Modern Architecture: A Critical History, 3d rev. ed. (London: Thames and Hudson, 1992), 307.

93. Thus, it is only in modern discourse that ornament (its function, history, and rules) becomes an independent topic of reflection, and what ornament is applied to does not necessarily enter the picture, in such writings as Owen Jones's influential The Grammar of Ornament (London: Day and Son, 1856); Alois Riegl, Problems of Style: Foundations for a History of Ornament (1893), trans. E. Kain (Princeton: Princeton University Press, 1992); Adolf Loos, "Ornament and Crime" (1908), in The Architecture of Adolf Loos, 2d ed. (London: Arts Council of Great Britain, 1987), 100-103; and, more recently, E. H. Gombrich, The Sense of Order (Oxford: Phaidon, 1979).

94. Alberti (as in n. 67), 163.

95. Ibid., 156.

96. Ibid., bks. 6-9. The pre- or nonmodern conception of ornament appears throughout these four books on the subject. For example, "Like the temple, the basilica should be set on a podium, but the height of the podium should be one eighth less, in keeping with its lower religious standing. All its other ornament should lack the gravity of that of a temple" (230); or "Watchtowers provide an excellent ornament, if sited in a suitable position and built on appropriate lines; if grouped closely together, they make an imposing sight from afar" (257). Nor is ornament - being paired with beauty in general, rather than specifically with architectural beauty - used to designate exclusively architectural qualifies or features: "The countryside along a route may be a considerable ornament to a military road, provided it is well maintained and cultivated, and full of villas and inns, and plenty of attractions; with views now of the sea, now of mountains, now of lakes, rivers or springs, now of parched rock or plain, and now of groves or valleys. If the road is neither steep, nor torturous, nor obstructed, but rolling as it were, level, and quite clear, it will also be an ornament. . . . Moreover, if the traveler often comes upon objects that stimulate conversation, especially if it is about high matters, that is an ornament of the greatest dignity" (244).

97. Ibid., 310.

98. Ibid., 61. The only thing Alberti excludes from structure are foundations: "The foundations, unless I am mistaken, are not part of the structure itself; rather they constitute a base on which the structure proper is to be raised and built."

99. Laugier, 129.

100. William Chambers, A Treatise on the Decorative Part of Civil Architecture, 3d ed. (1791; reprint, with a "Life of Sir William Chambers," and "An Examination of Grecian Architecture," by Joseph Gwilt, London: Priestly and Weale, 1825), 150; emphasis added.

101. "Ornament" could, however, also continue to have a more comprehensive meaning and signify an entire formal repertoire. Such a use of "ornament" is present in the treatise of J. L. de Cordemoy (upon which Laugier had heavily drawn for his Essai), as in the following passage, where gout (close to the modern "style") and ornemens seem to be virtually interchangeable: "Michelangelo truly brought honor upon himself for having revived the style [gout] of ancient architecture; but he would have done even better had he at the same time retained that which is good in Gothic architecture: I mean by this the spaciousness and sharp clarity of the intercolumnations which are so pleasing to us. For example, would not the churches of Royaumont, of Longpont, and of Sainte-Croix of Orleans be of the greatest beauty if they had the ornaments [ornemens] of ancient architecture?" Cordemoy, Nouveau traite de route l'architecture (Paris: Jean-Baptiste Coignard, 1714), 110. On the relationship between Laugier and Cordemoy, see Middleton, 26: 98-101.

102. Laugier, xvii.

103. Ibid., xvi.

104. Ibid., xvi-xviii.

105. Jennifer Bloomer has provided an incisive analysis of Alberti and the problem of the supplement in the distinction he seeks to make between beauty, which he tells us "is some inherent property," and ornament: "if the 'inherent property' is a sufficient condition for beauty, ornament, as an addition that for Alberti is a positive one ('Who would not claim to dwell more comfortably between walls that are ornate?. . .'), is in excess of the conditions for beauty, while at the same time pointing to a lack in the essentially beautiful (unornamented) object. There is the suggestion of the temporal condition here also: The beautiful object is first beautiful without ornament; ornament is added after the establishment of the beautiful object. When this occurs there must logically be a slipping away of beauty, because for the object to possess beauty in the first place, 'nothing may be added . . . but for the worse.' So when something (ornament) is added, the beautiful object becomes both worse (no longer its pure self) and better ('more delightful')"; Bloomer, "Tabbies of Bower," in Deconstruction and the Visual Arts: Art, Media, Architecture, ed. P. Brunette and D. Wills (Cambridge: Cambridge University Press, 1994), 231-32.

106. Alfred Barr, "Modern Architecture: International Exhibition: Foreword" (1932), in Defining Modern Art: Selected Writings of Alfred H. Barr, Jr., ed. A. Newman and I. Sandler (New York: Abrams, 1986), 79-80.

107. John Onians, Bearers of Meaning: The Classical Orders in Antiquity, the Middle Ages, and the Renaissance (Princeton: Princeton University Press, 1988), 153.

108. Some have seen the structure/ornament pair as already present in antiquity. For example, Thomas H. Beeby writes, "Until this century, ornament was conceived of as a primary aspect of architecture, and a building was constructed as a structural vessel to receive the veil of surface ornament. . . . In the case of classical architecture, evolving from Greece, through Rome and the Renaissance, to neoclassicism, the rules governing the placement and disposition of ornamental elements becomes quite elaborate"; Beeby, "The Grammar of Ornament/Ornament as Grammar," Via 3 (1977): 12.

109. Levine, 332.

110. Such sympathetic readings seek justification in the way their subjects had been perceived at the time they were writing. When Neil Levine undertook his reassessment of the Bibliotheque Ste-Genevieve, for example, he sought to challenge the prevailing modernist reading of the building that had been offered by Sigfried Giedion: "To preserve the building's prescient quality of utilitarian instrumentality, Giedion had to delete from his reconstruction of Labrouste's intention any mention of the decorative forms by which the structure of the library makes itself manifest as a work of art, in a word its clothing" (326). Consequently, a principal aim of Levine's study was to restore to the structure of the library the decorative "clothing" suppressed by Giedion. Yet we would do well to listen to Jacques Soulillou (as in n. 91), who writes about such attempts to rectify so-called modernist injustices, "Such a reactive interpretation only supports the illusion of an autonomous decorative space controlled by its own laws" (14).

111. Although this paper concerns structure/ornament and its relationship to nonnormative architecture, in fact, its narrative role has also been exploited to write about canonical buildings. To take one example: in his discussion "The Classical Age of Gothic Architecture," Louis Grodecki divides Reims Cathedral into its conceptual Chartres-derived model, or "structure," and its opulent ornamentation. The former allows the building to be located in the continuous formal history of Gothic architecture and to be narratable in systemic hypostatic term - as an improvement or critique on the Chartres type (itself a reconceptualization of earlier Gothic solutions), which will be subject to further mutations at Amiens and beyond. The lavish ornament (which neither hinders nor contributes to the progress of the Gothic) is explained through the exceptional purpose of the cathedral as a setting for the royal ceremony of the sacre, and is also understood as a regional architectural preference. The cathedral as a whole is interpreted as the synthetic product of a dialectical process between the specific contextual factors leading to the creation of this singular building and the overall evolution of Gothic architecture of which it is a moment. These two narrative lines are allowed to coexist and to be told simultaneously (and any conflict between them to be elided) because they are structured by the structure/ornament model; Grodecki, Gothic Architecture, trans. I. M. Paris (New York: Abrams, 1976), 119. On the problematics of "normative" architecture as distinct from "transitional," see n. 34 above.

112. The necessity of critical attention to this pair by contemporary architects is stressed in Kenneth Frampton's polemical rereading of the history of modern architecture, which takes the theoretical complexity of structure/ornament as a central theme. Frampton (strongly influenced by the writings of Gottfried Semper) characterizes structure and ornament in terms of the difference between the "ontological" and the "representational" aspects of architecture, which is a difference between "the core of the building that is simultaneously both its fundamental structure and its substance" and "the skin that re-presents the composite character of the construction"; Frampton (as in n. 82), 16.

113. A possible answer to the question, "If not structure/ornament, then what?" will be presented in my forthcoming book on St-Eustache.

Frequently Cited Sources

Bergdoll, Barry, Leon Vandoyer: Historicism in the Age of Industry (New York: Architectural History Foundation, 1994; distributed by MIT Press, Cambridge, Mass.).

Blunt, Anthony, Art and Architecture in France 1500-1700 (London: Penguin Books, 1953).

Laugier, Marc-Antoine, Essai sur l'architecture, 2d rev. ed. (Paris: Duchesne, 1755).

Lenoir, Albert, and Leon Vaudoyer, "Etudes d'architecture en France, ou notions relatives a lage et au style des monuments eleves a differentes epoques de notre histoire," Magasin pittoresque 7-20 (1839-52).

Levine, Neil, "The Romantic Idea of Architectural Legibility: Henri Labrouste and the Neo-Grec," in The Architecture of the Ecole des Beaux-Arts, ed. Arthur Drexler (New York: Museum of Modern Art, 1977; distributed by MIT Press, Cambridge, Mass.), 325-416.

Middleton, Robin, "The Abbe de Cordemoy and the Graeco-Gothic Ideal: A Prelude to Romantic Classicism," parts 1 and 2, Journal of the Warburg and Courtauld Institutes 25 (1962): 278-320; 26 (1963): 90-123.

Sauerlander, Willibald, "Style or Transition?: The Fallacies of Classification Discussed in the Light of German Architecture 1199-1260," Architectural History 30 (1987): 1-29.

Vachon, Marius, La Renaissance francaise: L'architecture nationale, les grands maitres macons (Paris: Flammarion, 1910).

Van Zanten, David, 1977, "Architectural Composition at the Ecole des Beaux-Arts from Charles Percier to Charles Garnier," in The Architecture of the Ecole des Beaux-Arts, 111-290.

-----, 1987, Designing Paris: The Architecture of Duban, Labrouste, Duc, and Vaudoyer (Cambridge, Mass.: MIT Press).

Viollet-le-Due, Eugene-Emmanuel, Dictionnaire raisonne de l'architecture francaise du XIe au XVIe siecle, vols. 1-6 (Paris: B. Bance, 1854-63), vols. 7-10 (Paris: A. Morel, 1864-68).

Anne-Marie Sankovitch, a research associate at the Institute of Fine Arts, New York University, has written about medieval and Renaissance architecture and theory. She is completing a book on the historical, historiographical, and theoretical issues attending the church of St-Eustache in Paris and the architecture of sixteenth-century France [Institute of Fine Arts, 1 East 78th St., New York, N.Y. 10021].

Browse by alphabet: