Early Christian and Byzantine Architecture characteristics, digests, interprets and presents in a con cise and beautiful style all the wealth of material accumulated on the Christian buildings until the Gothic war in the west and the fall of Constantinople in the east. ‘Ilie dividing line between early Christian an Byzantine architecture is drawn during the reign of Justinian I. The field of Krautheimer’s book was co ceived as the last, and indeed dragging on, phase late antique building rather chan, according to a m re common approach, an introduction to mediaeval wes ern architecture. It is dominated by the notion of the continuity of the Roman Empire, first Christianiz d, then definitely anchored in Byzantium. “Methodical choice or historical view emailed the disregar of practically all western architecture in the Ge manized parts of Europe until the Carolingian ren ratio In the first part is rehearsed the prehistory, as it were, of Christian architecture: loculi in the cata combs, community centers like the one at Doura-E ropos, shrines such as the memorial of St. Peter, he domus ecelesiae at S. Crisogono in Rome. Churc building on a vast scale was ushered in by the edic of Milan in 313.Order now
Part 2 covers first the Comta tinian churches in Rome and Jerusalem (those founded b Constantine in Constantinople arc examined in a fo lowing chapter), and then the religious buildin s in Constantinople, the new Rome, Jerusalem, the patr archal shape of Antioch, the cradle—or favourite ho e —of the cruciform martvrium, the new capitals in he west: Milan, Trier (with its outpost at Cologne) an Rome, where until the closing decades of the 4th ce – tury the pagan conservatism of the senatorial clas stood in the way and dimmed the brilliancy of churc architectural programs. The main new feature of th Constantinian basilica, the continuous transept, di not appear, as was believed until recently, in the Sav iour church of the Latcran, but at St Peter’s towar 324. K. holds that the transept was essentially a tra s versal mariyrium, a shrine, and may have bee fashioned after imperial palace architecture and erec ed above the memorial of St. Peter, counter o the theory that it was liturgirally brought about, in th west as well as in the cast More recently it has been reasserted that the continu ous transept achieves a tau plan symbolically con nected with the tau sign as the seal of salvation and a figure of the cross (E. Sauscr, in Lexicon fur Theolope und Kirehe For the first time the alpior mentioned by Eusebius as the “head” of the Martyrium on the Gol gorha is interpreted—and graphically reconstructed in fig. 16—as an open rotunda with its inner wreath of twelve columns, integrated within the chcvet of the Martyrium.
(A variant reconstitution would inter pose a sort of transept between the nave and the four aisles of the Martyrium, and, on the other hand, the place of the finding of the True Cross at the head of the Martyrium, after the plan of the basilica of Mar cellinus and Petrus in Rome and its connected mauso leum of Helena 1312-324).) One is strongly tempted to infer that Constantine had in mind the of the Golgotha Martyrium with its twelve columns, when he planned to be buried in the Apostolcion of Constantinople surrounded by two groups of six oryXai, symbolizing, as “i mage-col u m ns,” the twelve apostles. К, however, would locate the tomb of Con stantine, before its removal to a separate mausoleum after 357, at the very center of the cruciform church, directly under a central drum. A very important addition to the iconography of Early Christian architecture, presented by K. in CahArch it (1960) 15-40, concerns the huge funeral basilicac, or halls for the anniversary banquets, in the Campagna Romans: S. Sebastiano on the Via Appia, S. Agncsc on the Via Nomentana, SS. Mar cellinus and Petrus on the Via Labicana, S. Loreno fuori le mura. These vast, simple and functional struc tures provide the hitherto missing link between the underground martyrium and the tomh church. Two of them arc associated with imperial mausolea.
The 5th century (Part 3) is split between the eastern half of the Roman Empire which managed to he by passed by the Asian and Germanic invasions, and the Latin west which was progressively submerged by the barbarians. The great difficulty of distributing so much disparate material within a few geographical boundaries more specific than the so-called regional schools is mirrored in the association of Egypt, where architecture assumed a strong national and monastic flavor, with the Aegean coastlands, where Hellenistic characteristics lingered, and in the grouping of Syria with inland countries including not only Palestine and Jordan but the high plateau of Asia Minora bracketing by and large justified by the far reaching extension of the patriarchate of Antioch. North Afri ca (Cyrenaica apart) is squeezed, in the Iatin section, between Ravenna and, on the other hand. Southern Italy, Sicily and Spain, although the Algerian and Tunisian churches share more in common with the Egyptian ones of the cloisonne type and the basilicae of the Syrian hinterland than with anything in Ra venna or along the Tyrrhenian coastlands. In Egypt the date of the ruins of St. Menas in Abu Mina will possibly have to be shifted to 457 (p. 32, n. 29), in stead of being spread under the reign of Areadiut and Theodosius II (408-450).
The dates of the first churches in Egypt with a triconch transept such as Ilcrmopolis, or a triconch sanctuary like the White Monastery at Sohag, are left with an interroga tion mark: 430-40 (?) and ea. 440. An even later date would better account for the emergence and relative frequency in Egypt of the triconch transept, side by side with the trctoil martyrium along the basilica at Tc’bessa (not before 440), and with analogous plans before the palace of M’shatta: the F.piscopium at Rosra (ca. 51a?) and the palace at Kasr Ibn Wanlan. In terms of planning, the Greek churches are mainly characterized by a tripartite or a cross transept, a fea ture to which the author was instrumental in giving currency in two articles , and V Congrejso Ji Archcoiogia Cristiana a83fl). The type of those transepts, however, is not confined to Greece. It is met also in Egypt (Menas Basilica) and along the south coast of Asia Minor, and, one must add, in places as distant from each other as Gcncsarcth (Basilica of the Multiplication of the Loaves) and Tropaios in Bulgaria. It seems that the Greek (not exclusively Greek) transepts and the Roman or continuous transepts (neither exclusively Roman, vide S. Eusebio in Vercelli and St. Peter at Salona ManaStirine) stemmed from the same archi tectural concept and served analogous liturgical pur poses. Krauthcimer agrees that the preparation of the Eucharist and the reception of offerings took place in the wings of a cross or of a tripartite transept. But in Basilica at Perge the prothcsis and the diaconicon are located on both sides of the apse, sug gesting that the place of the clergy was reserved to the central area of the pseudo-cross transept and that the aisles continuing those of the nave were designed to screen the faithful from the service at the Ixma.
When aisles enclosed totally or partially a cross tran sept, their function, in the case of maior pilgrimage basilicas like the one of St. Menas, must have been to channel the traffic of the congregation. It is hard to admit that the plan of Perge and of other basilicas with cross or pseudo-cross transepts stands “in a tradi tion which assimilates the plan of Constantine’s church of the Holy Apostles into a basilica.” The derivation is on the contrary convincing in the cross church with aisle* at Gaza (401) followed by the Church of the Prophets, Apostles and Martyrs at Gerasa (465) and the 6th century cross church at Salona. But three (not very telling) examples would substantiate the origin of the Greek basilica with transept in Con stantinople (two in Ebcrsolt, notices 78, and the excavation in the 2 Scrai-courtyard. ArchAnt In the patriarchate of Antioch a group of churches is hallmarked by a double shell construction combined with a quatrefoil plan: the martvrium at Scleucia- Pieria. the cathedral (now recognized as such, formerly called the martyrium at R’safah, the cathedral at Rosra (plus the church of the Theotokos at Amida. so close in plan to the martyrium at Sclcocia-Picria). The quatrcfoil may represent a development of the funeral primitive cella truAora into a fella quad richora. The double shell may have appeared in the Golden Octagon of Constantine in Antioch, which had colonnaded aisles and, as a palatine chapel, was the very ancestor of SS. Sergios and Rakchos in Con- stantinople under Justinian [. Neither the quaucioil churches of the Antiochene type nor the Golden Octa gon were vaulted, contrary to the Roman triclinia, salutatoria and pavilions in palaces in which they arc supposed to have originated. Finally the problem of origin is obscured on the one hand by ihc enigmatic and incomplete tetraconch in the stoa of Hadrian in Athens, which the Bulgarian Red Church at Pcrustica resembles, and. on the other hand, by the double shell structure and quatrcfoil plan of S. Lorenzo in Milan.
Monncrct dc Villard recognized the Antiochene “air de famille” of S. Lorenzo, hut dated the church in the fich century (ОС 12 (1946) 374). followed by A. M. Schneider (Dolgcr and Schneider, whereas Krautheimer, taking sides with Chicrici, dates its beginning in 382. While churches of the martyrium type thrived in the East, the only important martyrium of the 5th century in Rome is S. Stefa no Rotondo. a simplified copy of the Anastas is rotunda on the Golgotha. But the Roman basilicas of the 5th century with their sense of proportion, the refinement of their profiles and the prcciousncss of their veneer decoration, belong to a classical revival the acme of which coincides with the reign of Sixtus IV, 432-440 (cf. the authors article in E. Panofsky’s Essays (1961) 291-302). True, the cultural tenor of Roman life remained pagan until Honorius and by and large rather indifferent to Christian art on a monumental scale. But the turning point could be pushed up into the reign of Pope Damasus (366-384), who introduced in iconography grandiose themes stating the primacy of Rome: the Tradilio Legit and the forerunner of Giotto’s Navicdla in the atrium of St. Peter.
In Salona, one of the most active centers of church building in the pcrfccturc of Italy, the double cathe dral is linked with the double one in Aquileia. the tradition of nearby Jstria and earlier examples (Trier). K., however, depends too exclusively on the regional thesis of the authors of Foru/iungen in Salona (1917 1939). The double cathedrals in Merovingian France and in Visigothic Spain establish the fact that double cathedrals were a general practice in the west, one church being a basilica for the congregation, the other a martyrium. thus reconciling two functions which, in the cast, were kept antithetic or isolated. In Ravenna the characteristics of the buildings under Galla Placidia point toward Milan, the previous capi tal (Santa Croce inherited the cruciform plan from the Apostoleion founded by St. Ambrose; blind ar cades give rhythm to the mausoleum attached to Santa Croce as at S. Simp’iciano of Milan) and also toward the eastern Mediterranean (polygonal apse of S. Giovanni Evangelista). S. Apollinarc Nuovo was not built by Thcodoric to honor St. Martin but. following the precedent of the Constantinian churches in Rome and Nicomedia. was dedicated to Christ under the title of Saviour.
Part 4 is made up of three chapters, the first one describing Hagia Sophia (a great achievement of evocation, blending the structural, aesthetic, liturgical and ceremonial aspects of the church) and buildings descended from palatine architecture: H. Sergios and Bakehos in Constantinople, or copying it: S. Vitale in Ravenna (of the latter the western features ate underlined). Chapter to is concerned with the cen tralization in a cross-in square, in a cross transept church or in a cruciform one, of the same key element of architectural design: the pendentive or sail dome. The cross-domed church may have been fully elaborated in the 5th century (H. David, Saloniki). On the other hand, domed basilicas in Cilicia (Mcriamlik. Alahan Monastir), with their short naves covered by timber roofs and their pyramidal timber dome before the chcvct. announce the new formula, translated into the new Byzantine technique of light vaults, of an eastern dome extending into longitudinal barrel vaulted hays at II. Irene in Constantinople and Pirdop in Rulgaria, in the last third of the 6th century.
Chap ter 11 treats of the architecture of the age of fustinian in the provinces (including those reconquered in the western Mediterranean area, Dalmatia and the pa triarchate of Grado), where the basilical type of building still steals the show as a western heritage, but integrates the chcvct features from south Asia Minor (S. Apollinarc in Classc in Ravenna, the cathe dral at Cari&n Grad), or is modified by the adjunc tion of barrel vaults (Bulgaria). The tctracooch cathe dral at Rsafah, for which there is only a date ante quem (553), cuts a lonely figure in that context. The juxtaposition of its plan with that of the rebuilt Church of the Nativity in Bethlehem (figs. 74-75) invites one to visualize the latter as the cross-breed of a basilican nave with four aisles and a syncopated tctraconch. rather than as the addition to a cross transept with aisles of the three axes of a triconch. In the introduction to Part 4, K. point out that the central plan was ideally suited to the require ment* of the liturgy as it became stabilized in the east from the 6th century on. The enveloping mas of the church encompasses, with the main dome as its center, the chancel connected with the amlo through a solca protruding more or less toward or into the nave, and the whole nave is henceforward used as a processional path for the little and great entrances.
The adaption of the central plan to the liturgy of the mass is striking indeed in buildings with a cross transept and a very short nave (Basilica В in Philippi and II. Titus in Gortyna). But to what extent docs it account for or did it contribute to prepare the main building to which that liturgical explana tion of Justinian architecture is applied, Hagia Sophia of Constantinople? The scheme at the core of Hagia Sophia, a cupola bracketed between two half-cupola on an cast-west axis, seems to have existed in the martyrium of St Stephen at Gaza, as described by Choricius in his iMudatio Mardani, and a similar one was excavated in the mausoleum of Mikulcice in Moravia (J. Cibulka. in Sancti CyriUi et Methodius Prague 19631 96-98). In terms of light symbolism and of the iconography of architecture, Hagia Sophia, particularly after the heightening of the dome after 557, and like the cathedral of Edcssa described in the Syrian sugitha (p. 160), was a cosmic house dominated by the dome of heaven, an archetype of the universe, the visible form of the divine source of light and the abode of God Pancreator. In tracing the transition between the dichotomy observed from the 4th to the 6th century between churches intended for the regular mass and martyria erected for the cult of the martyrs, and the adoption, under Justinian, of a tingle type of building, the dome»! centrally planned church, rooted in the martyrium but provided with a chcvct that evolved in the basilica (apse, forechoir, prothesis, diaconicoo), it is usual to bring out the progressive graft on the martyrium of sucn basilican elements of the sanctuary.
The platform or bema, found in the nave of a dozen and a half examples in Antiochene or Syrian churches during the intermediate period, might be considered as a stage in that fusion. That bema which E. Baldwin Smith interpreted in The Dome by the light of the TesUmentum Domini, would not have been an enclosure for the ambo but a place, screened o(T by veils, for the commemorutio “which priests and people offer with supplication The duplication of the main altar (by means of a raparftdwtanv in the place of commemoration’) is sug gested by the tradition of antiphonal singing in the patriarchate of Antioch. In an already late type of martyrium (Sclcucia-Pieria, Martyrium of St. Sergios at R’safah) the bema occupies a place which cor responds to that of the tomb and altar in the churches of the pure martyrium type (St. Babylas at Antioch Kaoussi£, St. John of Ephesos), equipoising the altar in the chevet. Finally, the bema in the nave was eliminated when the whole church, throughout the Fast, became shaped like a martyrium, but was ex clusively used for the synaxes and the divine liturgy of the mass celebrated at the altar—the cult of tnc martyrs and the saints being confined to a martyrium chapel attached to the church. The introduction to Part 5 and to Chapter 13 again deal with the key problem ol the origin and explora tion of the cross-domed church as evolved but dis tinct from the cross shaped church analyzed in Part. The church of the Koimesis in Nicaca (burnt down in 1922) could W the earliest example of this kind of domed church, since the mosaic of the stand ing Virgin in the apse, that replaced the cross set up by the iconoclasts, seem to have been preceded by a ргеiconodastic image of the 1 lodegetria. Theoretically, the first mosaic of the Virgin Hodegetria in the Koimesis church could date in the reign of the icono clast Leo III (717-740) who had this very type im pressed on some of his seals.
The church of Kasr ibn Wardan would point toward a Constantinopolitan prototype, because it was built with bricks imported from Constantinople. Rut its vault dome, heavy and of a narrow span, does not bespeak unadulterated Consiantinopolitan engineering. It seems to he the successor in brick construction of the wooden dome usual in Syria and its borderlands in connection with the inscribed cross plan. A compact domed basilica, like Kasr ibn Warden, presents all the elements found in H. Sophia of Salonica in the early 8th century: a naos planned as a cross inscribed in a square and surrounded on three sides by a continuous ambulatory and a triple chcvct. The triple sanctuary became canonical after the liturgical restructuring of the 6th century. Its general adoption makes one question whether in tnc cross-domed church the “mystery” of the mass was performed in the domed central bay. The symbolical boundary between naos and sanctuary was marked by the tcmplon screening off the chancel. But the location of the amtx> at the very center of the church, under the cupola, as at St. Nicholas of Myra for instance, could be explained by the symbolical value of the ambo in front of the altar where Christ is sacrificed, as the “stone rolled away from the sepulchre,” which the ministers of God ascend to an nounce the resurrection of Christ (Germanos. P.G. 98 col. 392; Simeon. P.G. 155. col. 345), so that the “dome of Heaven” above the church would have been a memorial of the Anastasis of Christ. In Chapter 14 connections arc delineated between Syria, Mesopotamia and the Tur Abdin on one hand, and Egypt on the other.
The Syrian type of the monophysitc church was imported into Egypt by the monastic congregation of the Tur Abdin. On an op posite axis of influences, architectural concepts born in Sassanian palaces and in Nestorian or monophysite congregations of Persia had a far reaching impact on the vaulted hall churches of Armenia and Bulgaria. In deed, many features of Bulgarian architecture would be diflicult to explain without positing the migration of forms created in the basilican churches of Armenia from the mid 6th century until the last third of the 7th century. No less manifest were counter-currents that led to a revival of the basilican type of the prc lustinian era at Aboba Pliska, or of Roman inausolca in the round church of Preslav, traditionally identified with the palatine chapel of the Bulgarian Czar Simeon (893-92)—a tradition sustained by the author. Armenia, as К. points out, was finally overcome by the great wave of ccntralizeJ building that radiated from Constantinople. The three plans juxtaposed on fig. 94 bring evidence against the one-sided Iranian thesis of f. Strzygowski. St. Hripsimc at Vagharshanat (618) is an adaptation of an early Christian quatrcfoil plan to four diagonal niches continuing to the pave ment the squinchcs of the cupola—St. Gayanf, also at Vagharshapat (630) docs not differ from a regular domed church, and only the plan of the church at Bagaran (630) may on paper recall an Iranian fire temple. In comparing the fire temples to the Christian churches in Persia, A. Upham Pope was careful to remind us that the reason for enclosing the four cen tral arches of a fire temple with a dark continuous Armenia, as К. points out, was finally overcome by the great wave of ccntralizeJ building that radiated from Constantinople. The three plans juxtaposed on 94 bring evidence against the one-sided Iranian thesis of f. Strzygowski. St. Hripsimc at Vagharshanat (618) is an adaptation of an early Christian quatrcfoil plan to four diagonal niches continuing to the pave ment the squinchcs of the cupola—St. Gayanf, also at Vagharshapat (630) docs not differ from a regular domed church, and only the plan of the church at Bagaran (630) may on paper recall an Iranian fire temple.
In comparing the fire temples to the Christian churches in Persia, A. Upham Pope was careful to remind us that the reason for enclosing the four cen tral arches of a fire temple with a dark continuous ambulatory was to prevent a ray of the sun from falling on the flame burning on the altar if Survey of Persian Art I, 5506). The light symbolism of a cupola over a centrally planned church aims at exactly the opposite. In more technical terms, S. Guycr re traced the byzantinization of the Sastanian squinch (Grundl. der rntuelalterlUhen abendlandischen Bju tyinst 171, 175). In spite of those reservations, Ar menian architecture is the only one to possess more than a provincial status and to stand with that of Constantinople “on an equal footing.” The country became the transmitter of the dome carried by corner squinches to Greece and Constantinople (p. 235). Part 6 presents the new types of building that were evolved in Constantinople and set the fashion for the regional schools, from the Macedonian Renaissance on, and under the houses of Ducas, Com menus and Angclus to the Ijtin conquest of 1204. The author opposes the propriety of the term “Renaissance” as applied to the architecture of the post-iconoclastic period, a term otherwise so fitting to describe the re newal of the humanities, the copy of Early Christian prototypes in book illumination and the flourishing of the decorative an in the X century.
What characterizes architecture during the “middle Byzantine Renais sance,” is a surprising dwarfing of scale of the monu ments, allied with inventiveness in the articulation of the structure and in interplay of volumes, and cer tainly not a deliberate return to architectural forms or concepts of the pre-iconoclastic era. Everybody who has had to struggle with the disquisitions of G. Millet in L’Ecole Grccque dans Гarchitecture byxantinc (1911) will remain thankful to Krautheimer for his clarifica tion of the chief families of churches: the atrophied Greek-cross plan, the octagon domed church, the Greek-cross octagon plan and the quincunx. The By zantine church that became standard under the Mace donian Dynasty: “a domed centre expanding into а cross and interpenetrating with an enveloping belt of subordinate spaces,” is linked to the principles of design that underlay the cross domed churches from Justinian to the early iconoclastic period. A funda mental change intervened, however, in the treatment of the inner masses and in the light effects. Whereas the pre-iconoclastic churches are sturdy and even heavy, sparsely lit and revetted with rather somber marble sheathings, those of the Macedonian era are commendable for their sense of intimacy and for the subtlety of their space-light relationships. One cannot help missing here quotations from the ckphrasis writ ten by the patriarch Fhotios on the Pharos church in Constantinople, where the new style appeared (864). That ckphrasis is a wonderful piece of baroque art criticism, anticipating a modern analysis in terms of Einfiihlung.
It evokes the dynamic and subjective ideal of Byantine architecture, already exemplified in Hagia Sophia, its surprising cross vistas, the psy chological transfer to the building, in terms of motion, of the emotion of the onlooker. K. would reserve the concept of “renaissance” to a few solitary, almost erratic buildings located in the western outposts of Byzantine architecture. The derivation of S. Marco in Venice from the Apostoleton in Constantinople is a locus classicus in mediaeval archaeology, except for the contention presented here that the cupolas of S. Marco followed a remodeling of Justinian’s Apoetolcion between 940 and 979. As for the cathedra] of Pisa, this monument cannot be con sidered as “almost a freak.” It is not Buschcto who began it in 1063. His name is not documented before 1104. It is not even established that he was responsible for the change of design that transformed (after 1087) a church, begun according to a paleochristian type with double aisles, into a cruciform church with transept arms of the martyrium family. The cathedral of Pisa is nearer to St. John of Ephcsos and to II. Irene in (‘-onsrantinoplc than to S. Dcmctrios of Saloniki. Part 7 deals with the agony of Byzantine archi- tecture under the Palcologues until the official death of Byzantium, more from the point of view of the art historian than from that of the archaeologist. It is a debunking of the praise lavished on the sophisticated and colorful church architecture of the decadence by G. Millet and Ch. Diehl. The new centers of vitality shifted towards the second empire of Bulgaria and the new Kingdom of Serbia. The masterly synthesis of Krautheimcr offers almost as many solutions as there arc problems raised by such a complex matter. Thanks to the author’s grasp of all the facets of each and every question, no issue is ever forced. His open mind constantly checked his science and allowed him to instill into so immense a survey light touches of irony, true to the Greek sprit of discussion by statement and denial.