On the origin of the early Christian basilica

A. Riegl, “Zur Entstehung der altchristlichen Basilika,” Jahrbuch der K.K. Zentral-Kommission für Erforschung und Erhaltung der Kunst- und Historischen Denkmale, NF 1 (1903), 195-216. Reprinted in K.M. Swoboda, ed., Gesammelte Aufsätze(Augsburg, 1928), 91-110.


Monuments of early Christian basilical building still stand in the territory of the Austrian Empire, indeed, some of the most perfect and the best-preserved; and if we furthermore consider the results of the excavations of the last decades, in particular those in Salona, then we will be able to venture the claim, without fear of contradiction, that today, at least within the West and outside of Italy, the Austrian monarchy can boast of the greatest number and the most important witnesses of early Christian basilical building. Thus it may be justified if with the following an attempt to solve the long-disputed question of the genesis of the early Christian basilica should be set down in these pages, which are devoted to the description of Austrian monuments of art.

That the Christian basilica at the time of its origin (roughly in the third century after Christ) could not have been conceived in isolation from the contemporary architecture of the Roman Empire – be it the profane or the pagan-religious – is clear already from the columned halls and the semicircular niches that were common to both. From this, one correctly concluded that the Christians of that era, when it came to the creation of a sacred building corresponding to the specific requirements of their cult, did not proceed to invent a new type, unknown before them, but made use of the forms available in the contemporary architecture of the Roman world-empire. However, the problem was then framed from the materialistic point of view, which dominated all art-historical research in the second half of the nineteenth century, as a purely utilitarian affair: the Christians had merely asked themselves, which of the building systems discovered by the pagans most readily accommodated the performance of the Christian cult, and in particular the ceremony of the Mass? So they identified the most suitable model for their own religious buildings – according to some, in the market basilica, according to others, in the private basilica, or indeed in the ancient villa complex writ large, etc. – and merely adapted that model according to the requirements of the new cult. For example, the oldest of these hypotheses of origin – and still the most fundamentally attractive, so long as one maintains the underlying assumptions – holds that the Christian basilica goes back to the model of the profane market basilica, and that the Christians altered its basic form only in one respect and for reasons of utility, namely, by removing the columns from the two short sides.

According to such an interpretation, however, which one may describe as purely antiquarian, something has been shortchanged: art. In the era of artistic materialism, this could be overlooked; for, according to that theory, the beautiful was indeed only a necessary consequence of the useful, the work of art merely a mechanical product of function, raw material, and technique. For the past decade, however, we have gradually learned again to distinguish more clearly between the useful, which corresponds to the satisfaction of sensible needs, and the beautiful, which pleases. But through the adoption of such a dualistic interpretation, the question had already arisen: what pleased the Christians in the basilica, so much so that they thoroughly preferred it from the very first and clung to it almost without exception over the course of the next millennium, at least in the West?

One could try to deny this question all validity from the outset, namely by arguing that the early Christians were at the very least indifferent, if not indeed hostile, to visual art. The earliest Christians were in fact full of eschatological notions; and even in later generations, when they came to terms with the world, and thus deigned to proceed with the construction of a church, we often still encounter among them ascetic inclinations which betray the lack of interest in, if not indeed an explicit aversion to, visual art. But it would be a mistake to overestimate the significance of such phenomena. Indeed, polytheistic paganism itself, which can in a certain sense be understood as the deification of visual art, produced spirits at its very core who denied to art all soteriological significance and thereby its right to exist: and in these ranks we encounter not only the neo-Platonists, whom one might perhaps judge as not entirely sufficient proof, since they belong to the later period; but also and above all the great Plato himself, who in many respects represents the classical summit of Hellenism. And indeed the rich funerary art of the primitive Christians should have prevented us from denying them a deeper inclination towards visual art. One need only compare the analogous relationships among us moderns, who in so many respects make a great effort to produce a formal cult of visual art, but in regard to the artistic treatment of burial grounds can regard ourselves as not so far distant from the early Christians. Therefore the early Christians, or at least the majority of them, felt a pronounced need for the creation of visual art and for relishful re-creation, and therefore emerged as fully justified co-creators of the art of the Roman world-empire, alongside the pagans, to whom alone the active role in this process was previously assigned. We are now faced with the question: what differentiates this presumably Christian Kunstwollen from the contemporary pagan Kunstwollen? How was their reciprocal relationship constituted?

An attempt to derive an answer directly from the monuments will not get very far, at least with the present state of research. Our journals, as already decades ago, still focus primarily on iconographic distinctions in the representations of figures; the suppression of objectionable pagan motifs presents itself most strikingly, but even this was not carried out with rigorous severity. If the iconographic comparison sticks more to the exterior, material qualities of the content, without pressing on to the interpretation, then we remain without insight into the distinctions in the treatment of form and color. But already this negative result alone is not without a certain explanatory significance: if the difference here were truly quite considerable and profound, then surely it would have struck a few researchers. We will also be led to the same realization [that is, of the essential similarity of the Christian and the pagan Kunstwollen] through considerations of a general nature.

It is our deeply rooted and inextinguishable conviction that an intrinsic connection exists between the ethical and the aesthetic Wollen. One expression of this conviction is the recurring temptation to establish connections between art-historical phenomena and those of contemporary cultural history, and to explain the former by means of the latter, as the more generally determined. The rigorous critics, who habitually find fault with such attempts, make easy sport of them by inevitably asking what might have linked the two phenomena. But this only serves to demonstrate the existence of a compelling and therefore entirely real drive, rather like the electrical current – if the comparison be permitted – which seeks the quickest route to its goal by the forcible means of the short circuit, in order to spare itself the detour over the usual, more circuitous wire. One will perhaps be allowed to advance the claim, that not one of these clever and cautious critics has not at one time or another sinned in the same fashion. It may even today already be possible to more precisely specify the intrinsic connections between the relationship of the early Christian Kunstwollen to the contemporary pagan Kunstwollen, on the other hand, and that of the early Christian religious Wollen to the contemporary pagan religious Wollen, on the other; for at least in the extreme cases, where the reciprocal connections escalate to the point that one side is negated, as during Iconoclasm, these connections [sc. between Kunstwollen and religious Wollen] often lie in plain sight. However, since such an attempt would be too circuitous for this venue, we must appeal to the inner, instinctive conviction of the reader, and begin by asking: how does the relationship of the early Christian religion to  contemporary pagan religion appear in the light of the newest research?

Until now, an impartial assessment of this relationship has come into conflict with the belief in revelation – not, admittedly, with that belief in itself, but on account of certain secondary conceptions which have been tied to it. Proceeding from the consideration that a message, which must have been bestowed from heaven, could not in any way have been discovered by men, one believed it necessary to close an unbridgeable gap between early Christianity and the paganism of the same time. The continual advance of historical knowledge, however, has gradually narrowed this gap, and the belief in revelation has suffered no damage thereby. Nor does belief in revelation contradict the assumption that the pagans in the Roman imperial era stove toward the same goals as the Christians; whereby the latter found in revelation the salvation that they desired, and which was denied to the pagans. The commonality of the fundamental religious and ethical goals of all peoples and cults in the Roman world-empire of the imperial period is now supported by too much evidence for us to keep our minds closed to it.

Growing recogntion of the inability of the traditional polytheistic faith to meet the ethical need for salvation felt by civilized humanity of the time: this characterizes religious life during the first three centuries of the Roman imperial era. Among the learned, this recognition had admittedly begun centuries before, when Greek philosophy sought to grasp that with the understanding, for which faith had failed. By the second century A.D., these classes of the population had arrived at skepsis, that is, at doubt in the possibility of salvation through philosophy and science. The great majority, however, could not live without an imperative belief, but having long since lost their faith in the cultic forms of the twelve-god pantheon of the Greeks, they grasped instead at new cults. The Roman of the imperial era did admittedly continue to venerate the old gods of state and family, rather as one fulfills a prescribed formality of citizenship; however, in the inner distress of his heart and his need for salvation, he no longer turned to Zeus or Hera, Apollo or Athena, but rather to Isis or Attis, Serapis or Mithras, and so on. All of these divinities, who in their external forms were borrowed from oriental cults, shared the following: first, a tendency towards monotheism; second, a ritually based mystery cult; and, finally, a connection to immortality, in which late antique man increasingly caught sight of true salvation. In all these connections, however, the opposition to the classical polytheistic outlook emerges just as clearly as the similarity to Christianity.

If, therefore, we may establish that Christians and pagans in the early imperial era shared a final, religious goal, whose attainment however theyb sought with differing means, then one might assume the same of the Kunstwollen, on the basis of the analogy outlined above. An example is not far to find, and offers itself in the basilica. That the Christian and pagan (legal) basilicas must have corresponded to closely related, if not identical, aesthetic demands, is indicated by the close correspondence of most of their parts. But a difference is also at hand: for no Christian basilica resembles the market basilica exactly, nor vice versa.


The foregoing considerations sought merely to clear away the obstacles that an earlier view, trapped in materialistic preconceptions, had placed in the way of the knowledge of the essence and origin of the Christian basilica. It had been alleged that the early Christians treated visual art rather like a tunic, which one put on or took off according to whether one wanted to be warm or cold, without acknowledging in the slightest its potential to awake the apprecation of the owner or of others. The most important result is that we are no longer compelled to seek a fixed building system among the pagans that the early Christians adopted ready-made and adapted to their practical ends. We may rather interpret and analyze the Christian basilica as a free artistic creation of the early Christians, albeit one made out of elements that were shared with the pagans, by virtue of a Kunstwollen that they also shared, at least it respect to its ultimate goals.

First of all, we must clarify the practical end that the Christian basilica had to serve; for even if the end does not, as the materialists had believed, mechanically produce beauty, it nevertheless provides the external impulse that brings beauty to life, and thereby determines its appearance at least in part. If one wishes to grasp purely that in the artwork which is to be attributed to the Kunstwollen, then one must be able to separate out of the total appearance that which is determined by the practical end; therefore the practical goal that the Christian cult building had to serve must be clearly perceived before beginning the artistic investigation. To this end, we may completely disregard the circumstances of the primitive era, which was still full of eschatological expectations, and which one may perhaps most fittingly designate as the communist era. The prerequisites for the construction of a generally binding type of religious structure were first present among the Christians in the moment when they began to prepare for a lasting stay in this earthly world until its unknown end. For various reasons, this change can be set no earlier than the second half of the second century A.D. This reversal carried with it simultaneously an internal consequence for the cult which, as will soon be shown, was also of consequence for the general formation of the religious structure.

The Christian religious structure, as it may gradually have developed since the end of the second century, was conceived as the site where the community participated in the saving benefits of the sacrifice of the mass, which was presented by the priests. And precisely this act was to be carried out behind closed doors: to ensure not secrecy, but purity, for no one ought witness the sacrifice, nor even glimpse it from afar, who was not internally prepared for it. In comparison to the sacrifices of the polytheistic cults, the Christian sacrifice had a thoroughly mystical significance, as also befit the contemporary Mithraic sacrifice and the like; and it was wholly consistent with this mystery that the believing individual, although not in fact excluded from bodily participation in the sacrifice, also participated in its saving graces when the sacrifice was merely presented by the priest in his role as the spiritual head of the community, and he – the believer – merely stood by, with a corresponding inner devotion, as the offering was made within the same enclosed space. However, this idea first took hold with the constution of a church; whereas during the communist era bodily participation in the sacrifice (its presentation and its acceptance) was obligatory for every believer. Therefore, with the change of approach, the priest, whose function in the communist era had been that of (if the not entirely adequate word be permitted) a negotiator, became far more elevated and desirous of respect in relation to the community. The Christian religious structure afforded therefore one space, in which the sacrifice was presented by the priest, and a second space, connected to the first but also respectfully separated from it, in which the community might gather; with both spaces protected from the gaze of the unbidden, those who did not belong to the community or who were unworthy to participate in the sacrificial graces.

One of the two spaces was the apse: half-cylindical, originally windowless, and covered by a half-dome; in which the mystery takes place, with the sacrificial altar and the priesthood. It is essentially one half of a rotunda with a round dome, and is therefore a descendant of the centrally planned building, perhaps the most characteristic motif in Roman imperial architecture. It is therefore necessary here to undertake a brief excursus on the meaning and significance of the Roman centrally planned building, granting that we know too little to speak with authority about its presumed Hellenistic ancestors.

The centrally planned building is a result of the emancipation of the dimension of depth in ancient art. This emancipation betrays itself not only in the exterior form of things, whose palpable exterior surfaces may now bend towards or away from the spectator, thus producing strong contrasts between light and shadow; but also by admitting the aesthetic potential of the open air. Herein lies its true epochal significance. Now, for the first time, after the classical phase of the development had been overcome, one began to perceive the interior space as an artistic element; from this moment on, it becomes possible to speak of an actual, monumental art of space. The challenge was therefore to fashion a balance between the intangible, infinite, formless depth of space, on the one hand; and the delimited, palpable, more or less symmetrically enclosed form, on the other. The solution lay in the centrally planned building. For example, upon entering the interior of the Roman Pantheon, one feels immediately that the depth of its space, which cannot be measured on the plane, and is therefore troubling, is roughly equal to the width and height, which can be measured on the plane. The round dome, by observing the same radius, balances out any uncomfortable angles, and thus completes the impression of certainty in the viewer. One senses the exterior, palpable, enclosed form also in the interior space, and forgets the immeasurable free space by virtue of the impression of the solid shape delimiting it. Symmetry form the basis both of classical art and of the centrally planned building; but whereas the former observes the symmetry of the plane, the latter observes symmetry of space, which since Semper has been called Eurhythmie.

The centrally planned building became the definitive monumental building of the Romans. This is demonstrated already by the fact that the tombs of the most distinguished -- monuments par excellence -- had been constructed as centrally planned buildings since the end of the Republican era. These are in a certain sense works of sculpture that are meant to present themselves to the outside world as enclosed, palpable, and therefore eurhythmically curved shapes. Yet as soon as the interiors were supplied with vaults and central plans, then a monumental significance dwelled also inside. In the ancient world, every sacral function possessed such a significance, and since the whole of ancient life was imbued with a sacral essence there were many such functions. Even spaces whose fundamental function was not monumental, and therefore conceived as simple and utilitarian, were supplied with little subsidiary spaces with some sort of sacral significance, and which thus exemplified the preferred artistic motif of the formed free space; in many cases it may have been the artistic need that supplied the actual inducement, the sacral significance only the pretext. Such subidiary spaces received of course not the complete central form, but its half: the half-circular niche crowned by a half-dome.

Thus it becomes clear, upon consideration of the Roman understanding of the significance of architectural forms, that the early Christians could only give to that an interior space distinguished by the presence both of the mystery and of its intermediary the form of the centrally planned building; and because it was necessary to add on a second space, communicating with the first, they could only choose its half-form, that is, the apse.[1]

The second space was conceived as the assembly room of the community. The purpose was, in comparison with that of the space of the mystery, decidedly less important. In general, the believers no longer had a direct role to play in the mystery itself; if one had distinguished their space likewise as monumental, then the priority of the space of the mystery would have been obscured. The space for the believers thus had to be treated like the usual assembly rooms of that time: spaces for the circulation of people, for coming and going, without any monumentally enclosed nature; in a word, columned halls.

The ancient hall (στοά, porticus) comprised a corridor perforated by a colonnade, where possible on all sides, but at the very least on one. The underlying thought was to create a delimited space for the assembly and circulation of people, while avoiding the impression of an enclosed interior space. Instead, the eye was directed toward the columns: delimited, palpable, material, individuals that stood in free space on level ground, like images in relief, and that suppressed the existence of the free space and rendered it invisible, just as the relief suppreses the ground below it. These halls were in a certain sense the ordinary type for assembly rooms in the Roman imperial era, as may best be seen in the market basilica.

The Roman market basilica was in its fundamental conception a composition of four colonnades that opened onto a shared, oblong court. In the interior, the colonnades formed the artistic element of the basilica: palpable, delimited, and material forms in rhythmic repetition. If the original conception required the court to be left open, it was soon roofed, for reasons that the ancients nowhere explain. The materialists would probably have recourse to practical utility to explain this momentous innovation, which would thus serve to keep the assembly room comfortable in summer and in winter, and in every kind of weather. Such an exterior motivation may indeed have played a role; but an artistic consideration must have run at least parallel with this practical consideration. Perhaps the same increasing drive towards spatial enclosure that took form in the cultivation of the centrally planned building was also manifest in the functional building. The peoples of the Mediterranean, who in the classical era sought the flat ground of free space as a necessary foil for the palpable individual, which alone possessed aesthetic validity, gradually began to feel a need for enclosure within a limited area. Put simply: they were embarrassed that unwanted observers could look into the market halls and observe their dealings from all sides (also from above), and thus not only did they close in the exterior sides of the hall with masonry walls, but they also shut in the court from above; and this in turn was tied to the erection of walls over the columns, in order to supply the requisite light. Thereby the hall ceased to be open, and this introduction of a closed interior surrounded by collonades became, if at first only latently, the point of departure for a new development, constructed on the basis of essentially unclassical conditions. But those who introduced this innovation (presumably already in the Hellenistic era) could not perceive the end of the development, which we today see so clearly. For the Romans of the imperial era, the roof of the court remained provisional, for it was never, so far as we know, constructed in the form of monumental vaulting, and not once in permanent material, but always as a flat roof of wooden beams.[2]

Now we will also understand why the early Christians gave the form of a three- (or more-) aisled naos to the assembly room that lay before the apse . The obligatory form for such spaces was the columned hall, and there was no reason for the early Christians to abandon it. The roofing of the central court was rendered all but necessary by the cult’s desire for enclosure from the outside. The disappearance of the colonnades on the short ends, through which the early Christian basilica is so strikingly distinguished from the market basilica, is thoroughly illuminating. There could be no direct connection between the space of the mystery and the space of the community, so that the space of the mystery had in a certain sense to give autonomously onto the open air, that is, onto the (roofed) court. On the other hand, the columns could well have remained on the opposite short end, as did happen here and there; but those halls that led to the mystery now gained a marked precedence, and therefore no real meaning remained for a transverse hall at the entrance. This was a crucial moment in the development of the type, in which a tendency already latent in the market basilica asserted itself more powerfully still in the Christian basilica. However, the new direction was certainly not articulated by means of a perspectivally conceived nave, rather only in the corridor-form of the side aisles and in the dominant relationship of their length to their breadth. Here too we must be careful not to ascribe those artistic intentions already to the early Christians, that later eras undertook to realize in the basilica.

Thus the two side aisles alone originally formed the constitutive element of the naos. This is not to say that the believers did not also enter the central aisle; so too in the atria, composed of an open court between four columned halls, will the members of the community have entered the court as often as the halls, although the masses were at least demonstrably excluded from a part of the central aisle – naturally the part that lay nearest to the altar (the schola cantorum). However that may be, the columns formed the primary artistic element of the early Christian basilica. Not the perspectival view from the nave towards the apse, but the view from a side-aisle, across the roofed court, and onto the façade of the other side-aisle, with its colonnade and paintings on the wall above: that determined the artistic effect of the naos in the early Christian basilica. Not one among the numerous early Christian basilicas in the West or the East that have survived in whole or in part possesses vaulting or even a coffered ceiling in stone: without exception, we find covering with a ceiling of wooden beams, with or without paneling, whose provisional character when viewed from an artistic standpoint so significantly contrasts with the half-dome of the monumental space of the mystery.

How, then, should we understand the relationship of the early Christian basilica to the market basilica? Was the latter the direct and conscious model for the former? We should not imagine that the early Christians made a choice after careful consideration of the available architectural types, which happened to fall on the market basilica. This becomes clear already from the conformity that prevailed across the entire Roman world-empire, wherever Christian communities constructed sacred buildings. Of course variations appeared here and there, indeed it would be strange had they not; but in general, within a short time the combination of the apse with two columned halls and a court inbetween was everywhere. This would certainly not have been possible if one place or another (even Rome) had chosen the market basilica as the model and the rest of the empire had then followed suit. No, the conformity may only be explained as follows: if the interpretation of architectural forms presented above was valid throughout the entire empire, so that the adoption of the same form for the same religious structure was a matter of necessity and required no additional impetus from the authorities. Such impetus would also poorly accord with the apparent reserve with which the church approached the visual arts during the first millennium.

Nevertheless, the relationship between the Christian and the market basilicas should not be ignored, for only thus can it be explained that the name ‘basilica’ became customary for churches. This relationship was based not only on the columned halls that surround or, as the case may be, flank a central court, but also on the apse. For the market basilica too had need of a monumental subsidiary space, which despite its integration into the whole was distinguished by means of its exceptional form: the judge’s tribune, which in its basic form corresponds completely with that of the apse, but in accordance with its lesser significance was much smaller, and certainly did not attract the reverence of all those assembled, serving instead the occasional interests of a portion. For this reason the columned hall suffered no interruption before it.[3]


When considering the view of the development of the early Christian basilica, in particular of its naos, developed above, many will at first glance find disconcerting the demotion of the nave from the rank of the primary aisle. For, according to this view, the genuinely active elements are the two side-aisles; the middle aisle is merely the ground of the relief, architectonically appraised a vacuum, a formless, empty space, only provisionally roofed. We are today so accustomed, owing to our own religious customs, to consider the nave as the actual space of the church, and the side-aisles merely as entries to the side-chapels, that we lack all sympathy for the reverse relationship. We assemble in the nave during the mass, we seek constantly to have an unobstructed view of the main altar and of the celebrating priests; even while we are reading in the prayer-book, we live in the reassuring knowledge that we may ascertain the stage of the sacrifice by means of a glance. And yet the early Christians are supposed to have huddled primarily in the side-aisles, so that perhaps only those standing under the arcades could observe with their own eyes the sacrifice of the mass!

Let us begin by considering the side of this objection that relates to the cult itself. Here one may mention the fact that eastern Christians are even today separated from altar and priest through the iconostasis; thus the unhindered view of altar and priest is no absolute condition of the Christian cult. This understanding seems to correspond very well to the character of the mystery, which bases salvation more on interior devotion than on exterior, sensible perceptions.[4] Furthermore the galleries, which are themselves also side-aisles, attest that those very spaces, which according to the modern conception are subsidiary, served to accommodate the believers. That the galleries served to accommodate the members of one sex makes it at least highly probable that the two side-halls of the ground floor were also originally, and for a long time thereafter, tied up with the separation of the sexes.

What interests us here, however, is primarily the artistic side of the question. Today we fully enjoy, consciously or unconsciously, the spatial effect of the nave as an enclosed spatial whole; and we also enjoy its perspectival effect, insofar as it leads our glance to the main altar as a goal. It seems quite obvious to us that the early Christians should already have sought and enjoyed the same effect. However, a historical inquiry will not bear out this preconception.

Still today I vividly recall the hours that I spent in student years at the entrances of San Paolo fuori le mura and of Santa Maria Maggorie, seeking fully to experience the artistic effect of their naves, and thence to draw conclusions about the artistic intention of their erstwhile authors. Naturally I too sought a spatial and a perspectival effect. In San Paolo fuori le mura (which, even if it is only a copy, should be considered a valid record of spatial relationships), the perspectival effect appeared to be present at least in one passage: the arcades, which, as all the handbooks explained, led the viewer’s gaze to the altar. But why, then, the vast walls over the small, graceful arcades, and why indeed the enormous, immovable roof? In vain did I seek to chase off these misgivings with thoughts of “barbarism” and the Chrstians’ delibrate “hatred of art”: for the absurdity and untruth of such prevarications sat uneasily with me as soon as I had set foot in other interiors of the later imperial era, such as Santa Maria degli Angeli and the Basilica of Maxentius. In Santa Maria Maggiore, by contrast, even the arcades fell short. And strange to relate: if I placed myself roughly in the center of the nave, then the strange and the burdensome gradually receded, the columns achieved their effect, the solid, perforated upper walls appeared less cumbersome and seemed even, in the alternation of wall and perforation, to exhibit a certain harmony with the columns and intercolumniations below. Even the roof lost its rigidity and ponderousness, perhaps because from this standpoint it imposed itself less on the eye. Already these observations might have led me to seek the artistic efficacy, not in the perspectival effect of an enclosed space, but in the flat facades of the side aisles. But the prejudice in favor of the nave as the presumed bearer of the fundamental artistic intention was so great that I terminated my studies at the time with the discouraging confession that I had sought in vain the solution of the problem.[5]

Enlightenment should not have eluded me for long. The iconostasis, already mentioned, should make it clear that no perspectival view of the altar was intended; so too the realization that the chancel plaques, ambos, Easter candles, etc. must already have obstructed the view from the nave to the altar. The Plan of Saint Gall, by showing the entire nave full of altars, gestures in this direction; that is, if one may draw conclusions about the organization of a congregational church from that of a monastic church. What is more, the era [of late antiquity] displayed no understanding of perspectival effects, a point born out by the entire historical development of medieval art. What one would impute to the early Christians, by attributing a perspectival effect to the naves of their churches, can first be shown fully unfolded in the fifteenth century: in the churches of Brunelleschi, the reliefs of Donatello, and the paintings of Jan van Eyck. How would one treat an enclosed, axial interior with perspectival intent? This can be learned from Baroque architecture, beginning with Giacomo della Porta – from the same Baroque style that removed those fixtures from the early Christian basilica that inhibited the perspectival view; and, moreover, walled in most of the smaller windows of its upper walls, and in their place set a few larger ones, in order to acheive a painterly effect. The nave of the early Christian basilica offered neither an enclosed spatial impression nor a perspectival view of the altar. What then did it really mean?

Five years ago, when I wrote the passage in the “Spätrömische Kunstindustrie” that touched upon this question, I could already express the conviction that the nave of the legal basilica must once have been an open court, a foil for the columned halls that ran about it. And yet modern practice and views still held such sway, that I myself did not yet dare at that time to draw the final consequence and to eliminate the nave altogether from the list of artistic factors. In explicit contradiction to the insight that I had expressed, I myself continued to pose the question, what other artistic intentions might have led to the formation of the nave besides spatial effect and perspective. It seemed possible to utilize the observation that imperial Roman architecture also delimited spaces of unambiguous enclosure, namely the centrally planned spaces, by means not of curving walls, but of a sort of free spatial sphere, analogus to the shadows at the edges of late antique reliefs. This can be envisioned by means of the developmental series Pantheon – Minerva Medica – Santa Costanza. I thought that the opening of the walls of the nave through intercolumniation had emerged from a similar intention, namely to render sensible the free space that circulated behind. But this can explain neither the divergent treatment of the short ends, nor the roof, and thus the hypothesis already then seemed unsatisfactory to me. Since then, I have abandoned it entirely, among other reasons because it renders incomprehensible the doubling of the side-aisles, a feature of the largest and the most monumental of the basilicas. If, however, one perceives the side-aisles as the constitutive elements, then their multiplication appears as the easiest and most obvious means by which to render legible larger assembly spaces to especially populous communities; if the nave had served the purpose of assembly, then it would have sufficed to extend its length and breadth. And what holds for the multiplication of side-aisles in width, holds too for their multiplication in height, that is, for the galleries.

The emancipation of the nave from the subsidiary function of a foil or relief-ground begins when the schola cantorum advances from the apse into the nave, continues with the increasing clutter of altars that we see in the Plan of Saint Gall, and culminates with the introduction of multiple choirs. But the nave was not truly conceived as an enclosed interior until it was vaulted: vaulting is indeed nothing other than the monumental expression of that transformation in the interpretation of the Christian sacred building. The space of the community now gains monumental significance equal to that of the space of the altar, which had been vaulted from the very beginning. The means is closely related to that which that the Romans had used before: the naos was divided into square bays with cross-vaults, which hold a balance between centrally-planned space and axial space. Explicit directionality is known in the middle ages only on the axis of height, not of length.[6] With the Gothic, the nave steadily increases in significance, while the side-aisles become rows of chapels. The art of the Baroque draws the consequences. The church of the Roman Baroque knows only a single enclosed and vaulted hall with the space of the altar as its perspectival conclusion. The side aisles have disappeared and chapels have appeared in their place, much as happened before in Roman baths. It is no accident that Michelangelo constructed just such a hall at Santa Maria degli Angeli; but his disciples replaced his cross-vaults with barrel vaults running in a single direction.

Thus in the West was the space of the community gradually emancipated; but in the Christian East it happened much earlier. It is true that the bipartite basilica was the ordinary church type here as well; this is shown above all by the Syrian monuments first published by de Vogué. But the caesaropapist tendency, which eastern Christendom had embraced from the start, did not allow the distinction between spiritual and secular drawn as strictly as in the West.[7] Therefore, in the East, the space of the community soon flowed together with that of the sanctuary, insofar as both were situated within a single centrally-planned building. But if the Orientals set a lower barrier between priests and communities, how much greater was the barrier between man and God; therefore, in the Orient, it was the site on which the divinity revealed itself, the altar, that remained hidden from the eyes of the multitude by the iconostasis. The internal relationship between the Byzantine centrally planned church and caesaropapism had long been acknowledged; in the foregoing the aspect has been revealed, from which the connection, to this point merely suspected, can be demonstrated through a series of concrete links.

[1] The regular erection of baptisteries as centrally-planned buildings also relates, among other factors, to the reservation of the centrally-planned building for interior spaces of monumental character. Baptisteries were themselves monumental frames for the piscinae, that is, for the solemn spaces in which the admission into the Christian community took place. And no one will be surprised that funerary monuments and funerary churches should have followed the same schema.

[2] The vaults of the Basilica of Maxentius on the Roman Forum are in this sense a unicum. The ground plan also deviates from the typical basilical scheme. It should be understood as the rudimentary stage of a development that was never taken any further, just as the religious policy of Diocletian and like-minded folk found its abrupt end, without the slightest trace of a follower, in the same Maxentius. Numerous monuments show that Romans knew how to vault oblong constructions just as well as the centrally planned. However, according to the evidence of the monuments (namely of the Roman bathing halls), they preferred to divide monumental axial spaces into central spaces, that is, at least into square bays with cross-vaults above.

[3] It will now be clear why the question of the specific region in which the early Christian basilica could first have gained its particular form was not at all raised in the preceding discussion. Columned halls and apses, in a form originally created by the Greeks, are to be seen throughout the Roman world-empire; and if one may speak of a Roman world-art, then it would certainly include these fundamental elements of Roman architecture. But the question of “Orient or Rome,” which is today so temperamentally debated, appears entirely indifferent from the standpoint of the research which the present remarks are meant to honor. How does it advance our knowledge, should we decide – as most presumptuously demanded – to replace the general designation “late Roman” with “Oriental,” even though no one could claim that the art of the fourth through the eighth centuries A.D. was simply identical with that of the ancient Egyptians, Mesopotamians, or Persians? How would it help us even to know whether a motif was used earlier in the east or in the west of the world-empire, if we do not understand it on the basis of its artistic significance and of the origins of this significance? And we are in general still very far from such an understanding of early medieval art. It has for decades been the ambition of the present author to pave the way for this understanding, and he will not let himself be led astray from the persistence and assurance with which he pursues the goal he has in mind, namely, of a true historical understanding of late antique art: neither through the continual passionate attacks of certain researchers, nor through the neutrality of the rest.

[4] For this reason, the explanation attempted by Felix Witting (Die Anfänge christlicher Architektur) also founders, despite some solid fundamental thoughts and useful observations, because it is based on the assumption of a “visual impression” which is suggested by modern impressions and not capable of historical proof. {Felix Witting, Die Anfänge christlicher Architektur: Gedanken über Wesen und Entstehung der christlichen Basilika. Zur Kunstgeschichte des Auslandes 10. (Strassburg, 1902).}

[5] The personal nature of this description will perhaps be forgiven if one considers that the difficulties attendant upon placing oneself today in the mindset of the late antique world may thus most vividly and convincingly be explained.

[6] The ambiguous statement in the Spätrömischen Kunstindustrie, p. 29, according to which Roman art had already consciously aimed at perspectival effects, should thus be emended.

[7] The transept is also a manifestation of the strict differentiation between monumental space of the mystery and profane space of the community; it clearly serves no other purpose than to differentiate more firmly between apse and naos, and its implementation was characteristically limited to the West. – As proof of the validity of the derivation of the basilical plan offered in this essay, I believe I might finally be able to adduce, in a certain respect, the oldest mosque plans as well, for example, the mosques of ‘Amr and of Ibn Tulun in Cairo.

-- Translation and {notes} November 2007 / May 2020