image index
home page
Christian Architecture

Click images to enlarge.
Italy, Milan, San Lorenzo
Professor Dale Kinney

San Lorenzo in Milan is an enigma. It is a daring building on a grand scale, much admired by Renaissance architects and especially by Bramante, whose design for the dome of St. Peter's shows its influence, as does the closely related pilgrimage church of Santa Maria della Consolazione at Todi (1508). Yet modern architectural historians tend to disparage it as not-quite-great. Despite the fact that it may be the earliest known example of its type (the "aisled tetraconch"), W.E. Kleinbauer calls it "a derivative building". Moreover, no one in late antiquity wrote its history, so we have no documentary evidence of when it was built, why, or for whom. The origins of San Lorenzo remain one of the great unsolved mysteries of early Christian architecture.

San Lorenzo is the centerpiece of a complex of buildings erected outside the Roman wall of Milan - though well inside the modern city - aligned with the ancient road to Pavia. It was separated from the road by an atrium fronted by a grand colonnade of sixteen fluted Corinthian columns, second-century in style and therefore reused from an earlier building. Smaller buildings are attached to the main church on the east-west and north-south axes: the small niched octagon of San Sisto on the north; Sant'Ippolito, a larger octagon with a cruciform interior, on the east; and the still larger octagon of Sant'Aquilino, identical in plan to San Sisto, on the south.

San Lorenzo itself has a symmetrical, four-lobed shape called a "tetraconch", from the Greek words for "four" and "shell [conch]". Geometrically it is a square whose sides curve outward to make the four "lobes". In San Lorenzo, the corners of the square are exaggerated to become towers, which appear on the plan as four small squares between the curved walls. The four-lobed perimeter is echoed internally by curved colonnades anchored by four piers. This type of design, in which the outer wall is doubled inside by another enclosure, is called a "double-shell". Double-shell designs seem to have been a late antique invention; the first clear example is Santa Costanza in Rome, built as a mausoleum in the middle or second half of the fourth century. Santa Costanza has a circular outer wall and a circular interior colonnade. Double-shells are also frequently octagonal, like San Vitale in Ravenna of the sixth century. Double-shells flourished in the fourth through sixth centuries, then became rare. They are in the genealogy of the unique design of Hagia Sophia, which maximizes double-shell principles and applies them to the longitudinal layout of a basilica.

The double shell design moves the load-bearing function of the outer walls to the inner "shell" of supports. This favors structural efficiency by decreasing the span of the dome or vault over the center, and making it possible to concentrate the weight of the superstructure on a few relatively massive supports (the four piers in San Lorenzo), which can be buttressed if necessary in the space between inner and outer shells. In San Vitale, the eight piers that mark the inner octagon carry the dome by themselves, but in San Lorenzo the four piers that carry the superstructure are joined to four towers in the spaces between the four "conches" of the plan. Depending on the form of the vault, the towers may have had a buttressing function, in addition to housing the stairwells that provided transit between the first and second levels of the ambulatory. Today the central space is covered by a tall ribbed dome that was designed by the architect Martino Bassi after a prior dome collapsed in 1573. It dominates the exterior view of San Lorenzo and gives it a verticality that is probably quite unlike the original silhouette.

Below the dome, the exterior still shows much original fabric. The original masonry employs the characteristic Lombard brick, thicker and shorter than bricks used in Rome, and deep red in color. The thin-walled perimeter of the tetraconch probably is of this masonry, now covered by stucco, and it is plainly visible on the towers and chapels. The best-preserved tower is the northeast one, which retains most of its early Christian fenestration; the others have been cut down (northwest), given typical Romanesque slit windows (southeast), or conversely, opened up at the top with a loggia for bells (southwest). The best-preserved chapel is Sant'Aquilino, which is unchanged including its dome, except for the exterior "dwarf gallery", where the original columns have been replaced by thin brick piers.

The interior elevation of San Lorenzo does not survive except in outline. The stone piers at the corners of the central space are connected by two-storied colonnades with ambulatories behind them. No two of these two-storied colonnades are alike; some have masonry columns or piers; one has monolithic columns; most are arcaded but one has a horizontal trabeation, etc. The general appearance of the inner shell comes from the late 16th century, but the colonnades may have been partially rebuilt before then as well. We might imagine that originally all of the colonnades on both stories were composed of monolithic marble columns and marble capitals (probably all spolia) and that all were arcaded, like the interior of San Vitale. San Lorenzo had much broader proportions than San Vitale, however, and a more expansive effect, not only because it was four-sided rather than octagonal but because it is significantly larger. Its north-south clear width is 47.9 m. The interior square is 24 m. on a side, and also 24 m. high.

The arcades must have been damaged in the major fires that devastated Milan in the middle ages and in the two known collapses of its dome. Two fires are documented between 1070 and 1075, and another in 1123 or 1124. In this period there was unquestionably some rebuilding. A drawing made ca. 1573 purports to show the dome as it existed before the collapse of that year; there has been some debate about whether it shows the original vaulting, repaired in the 11th-12th centuries, or a totally Romanesque replacement. A dome 25 m. in diameter would have been a major feat for Romanesque builders; on the other hand, it is questionable that the original late Roman designers would have projected a dome over this square base. The presence of corner towers has suggested to some that the original covering was a masonry groined vault springing from the four corners; still others have proposed a low wooden tower.

In the absence of documentation, historians have looked to circumstantial evidence to date San Lorenzo. A massive stone platform (platea) under the vestibule and part of the octagon of Sant'Aquilino contains molded and profiled blocks which seem to have come from a partial dismantling of the nearby Roman amphitheater. Thus for some scholars, an event alleged to have occurred in the amphitheater in 396 marks the earliest possible date for San Lorenzo; others point to an edict prohibiting gladiatorial combats of 405, which would have made the amphitheater obsolete; still others point to the removal of the imperial court from Milan to Ravenna in 408. San Lorenzo's grandiose colonnaded entrance, the scale and complexity of its design, and the probable funerary use of the smaller buildings attached to it make imperial patronage more plausible than construction by a bishop, especially in the later fifth century when Milan was beset by repeated incursions of "barbarians".

The fact remains, however, that we do not know who built San Lorenzo, or for what purpose. Although I myself have argued for a dating in the third quarter of the fourth century, I no longer believe that it is possible to be so conclusive. There is evidence for and against every possible date between ca. 350 and ca. 410. The architect Roberto Cecchi drew attention to an important consideration that others have overlooked, namely the use of ribs of hollow clay tubes (tubi fittili) in the vault of Sant' Aquilino. Tubi fittili were used to much better effect in Ravenna, where a dome of clay tubes was daringly inserted over the Orthodox Baptistery ca. 450, and in the larger dome of San Vitale ca. 540. Cecchi infers that the San Lorenzo complex represents a prior phase of experimentation with this innovative structural material, which he would date around 400.

We are equally unclear about San Lorenzo's purpose and function. Even the original dedication is not known for sure, although the dedication to St. Lawrence (=Lorenzo) does go back at least to the fifth century, when the cult of this Roman deacon (martyred in the third century) was very popular. It is not known how the original building functioned liturgically, or where the altar stood. Today the altar is on a platform in front of the eastern interior colonnade, but there is no good evidence that it stood there originally, or even that there was an altar at all. The location of the complex outside the city walls is evidence that it had a funerary purpose (since burial within the walls was forbidden), and the small chapel of San Sisto definitely was built as a tomb by Bishop Lawrence I (d. ca. 500). San Sisto is a later addition to San Lorenzo, as indicated by the fact that its masonry is not in bond with the main building. Sant'Aquilino, which is original, has the shape of contemporary mausolea and baptisteries. Most scholars think that Sant' Aquilino was meant to be the tomb of the founder or founders of San Lorenzo, whoever they might have been.

San Lorenzo is the only fully standing example of the double-shell, or aisled tetraconch. Twenty-one other examples have been identified in ruins, including one in Italy, at Canosa in Apulia. The others are in Greece, the Balkans, Egypt, Syria, Armenia, and ancient Mesopotamia. None of these ruins seems to date earlier than ca. 400 CE. A number of scholars have tried to identify the origin of the tetraconch as a type (Kleinbauer, using terminology first applied by Richard Krautheimer to the early Christian basilica, calls it a genus); usually the postulated prototype is Constantinian. One hypothesis is that the first tetraconch was the cathedral of Antioch, destroyed in the early middle ages, which is referred to in certain sources as the "Golden Octagon". Kleinbauer, who once supported this hypothesis, now argues instead that the prototype was a church in Constantinople. All such proposals are speculative and incapable of proof one way or the other. It might be more fruitful to focus on the distribution and function of the known examples; for example, Kleinbauer's recent study of the tetraconch in Athens highlights the probable civic prominence of this substantial building erected around the same time as San Lorenzo, in the repaired court of the Library of Hadrian.

One of the most pleasing effects of the double-shell design is to diversify the interior domed or vaulted space. Compared to the undifferentiated expanse under the dome of the Pantheon, for example, the interior of San Lorenzo offers contrasting experiences of open continuous space in the center and bounded, contoured and interrupted spaces behind the internal colonnades. When the enveloping spaces between outer and inner shells are double-storied the design also offers multiple elevations, as in San Lorenzo, with its gallery above the ground-level ambulatory. Multiple elevations create multiple and diverse viewing positions. While in the Pantheon the only viewpoints are from within the domed center and directed upward, San Lorenzo offers views into the center space from outside it, downward from the gallery as well as upward from the ambulatory. The centralized double-shell is thus a highly spectatorial space, and also an auditorial one. The design brings more people closer to the presumed focal point than the longitudinal basilica, making it possible for more people to see and hear clearly whatever is happening there. Nevertheless, it was the basilica that won out in the west as the standard form of church building, perhaps because it was easier to build, and perhaps also precisely because of the tetraconch's relatively more egalitarian character, which was counter to the tendency to make the Christian liturgy progressively more hierarchical and hieratic.


W. Eugene Kleinbauer, "Christian or Secular? The Tetraconch in the So-called Library of Hadrian at Athens", in Nova Doctrina Vetusque. Essays on Early Christianity in Honor of Fredric W. Schlatter, S.J., eds. Douglas Kries and Catherine Brown Tkacz (New York: Peter Lang, 1999), 203–224.

Dale Kinney, "Le chiese paleocristiane di Mediolanum", in Il Millennio ambrosiano. Milano, una capitale da Ambrogio ai Carolingi, ed. Carlo Bertelli (Milan: Electa, 1987).

W. Eugene Kleinbauer, "The Double-Shell Tetraconch Building at Perge in Pamphylia and the Origin of the Architectural Genus", Dumbarton Oaks Papers 41 (1987), 277–293.

La Basilica di San Lorenzo in Milano, ed. Gian Alberto Dell'Acqua (Milan: Banca Popolare di Milano, 1985), esp. Roberto Cecchi, "Architettura. Alcuni momenti costruttivi", pp. 79–116.

Richard Krautheimer, Three Christian Capitals. Topography and Politics (Berkeley & Los Angeles: University of California Press, 1983), 81-92.

Suzanne Lewis, "S. Lorenzo Revisited: A Theodosian Palace Church at Milan", Journal of the Society of Architectural Historians 32 (1973), 197–222.

Dale Kinney, "The Evidence for the Dating of San Lorenzo in Milan", Journal of the Society of Architectural Historians 31 (1972), 92–107.

W.E. Kleinbauer, "Toward a Dating of San Lorenzo in Milan", Arte lombarda 13 (1968), 1–19.

back to top

briefing | image index | resources
home page | site image index | site resources

media center for art history, archaeology and historic preservationcolumbia university