Clues from a Latin biography, a dozen Roman palace walls, and a nonexistent saint have solved one of the older riddles in European scholarship, the origins of the church that once stood beneath Florence’s cathedral. Taken together, the biography and decades of archaeological excavation and interpretation explain the origins of the Duomo of Santa Maria del Fiore. Unraveling this riddle tells us a good deal about how medieval culture worked, and how contemporary scholarship can still open doors that were once regarded as sealed forever.

During my direction of the Florence Cathedral excavation and after, I have studied the Duomo and used archaeology to draw a sketch of the earliest years of medieval Florence -- years for which documentary history is almost completely silent. Solving the mystery of Saint Reparata’s church, the ruins of which lie beneath the Duomo, represents a fragment of my work, but it illustrates the methods and reasoning necessary in studying medieval archaeology.

The hardest of my tasks was to date and theoretically reconstruct the St. Reparata ruins, determine why the Florentines had picked that particular building site, and why they had incorporated large chunks of a Roman palace into the building walls. I also had to figure out why the Florentines dedicated the church to Reparata, an obscure -- in fact, nonexistent -- martyr with no clear connection to their city.

The nonexistence of St. Reparata became an acute political issue in 1352, when Naples played one of the crueler tricks of medieval history on Florence. Florence, strangely, and perhaps tellingly, claimed no relic of Reparata even though she was patron of its cathedral. But a town in Neapolitan territory did. Legends related that the body of Reparata, who had been martyred as a teenager in A.D. 250 at Caesarea in what is now Israel, had floated across the Mediterranean to a port north of Naples around six hundred years later.

The Florentines asked the Neapolitans for Reparata’s right arm, which they reverently carried in procession through the city the next year. But when the Florentines unwrapped the arm for placement in a reliquary, they found that Naples had sent them nothing more than a plaster cast. Whether this deception stemmed from the Neapolitans’ reverence for the body of Reparata or, conversely, from their own doubts about its authenticity, we shall never know.

Naples’s treachery pushed the Council of the Republic of Florence to hold a special session in October 1353. Its aim was to discover who Reparata was and why the Florence Cathedral had been built in her honor centuries before. Finding no documents, the Council was left with the old and unreliable legend that Reparata had engineered Florence’s victory against some two hundred thousand Germanic troops who had besieged the city in 406. The victory of 406 is a historical fact, cited even by St. Augustine in his City of God, but Reparata’s participation in the events is so improbable that even in the Middle Ages few Florentines appear to have believed in it.

In consequence, Florence declared that the Virgin Mary, and not Reparata, would be the patron saint of the new cathedral it was beginning to build as a replacement for St. Reparata’s church. In 1412 the Republic forbade citizens to even mutter the name Reparata as the cathedral’s title-saint. When the Duomo opened a generation later, not one of its sixteen altars was dedicated to her.

What is striking in this tale of St. Reparata is how little the Florentines of the Late Middle Ages knew of their own early medieval history. By the time of Dante, even though the church of St. Reparata was then still standing, the Florentines had no clear-cut idea of its age, its original name, the patron who built it, or even why it was built. In contrast, the earlier and far better-documented ex-cathedral of St. Lorenzo stood just a few blocks away.

The last parts of St. Reparata were torn down around 1375. It came to light again in 1965 in the middle of the nave of Santa Maria del Fiore, with some of its walls just centimeters below the Duomo’s marble floor. The imposing dimensions of the earlier church, the richness of its mosaic floor, and its numerous structural ties to an underlying Roman domus, or palace, were taken by Florentine scholars as signs that the church dated from the fourth or the early fifth century. These vague indications of an “early” date seduced the same scholars into assuming that the victory legend of 406 was literally true: they reasoned that the church had been built right after that victory, to thank Reparata for her help.

But when I took over the excavation in 1969, I came to a set of totally different conclusions. I argued on the basis of numismatic evidence, radiocarbon dating, and a close reading of stratigraphy -- all three methods newly introduced to the excavation -- that the domus below St. Reparata was still standing, as a house, until at least the late fifth century. That meant that the church could not have materialized on the site until about a century after the victory date of 406.

My late dating of the church stemmed also from my analysis of the floor mosaic, which on stylistic bases compared to dated floors of the late fifth or early sixth century. In 1975, I proposed that the construction of the church of St. Reparata probably came around A.D. 500, during a break in the deteriorating military situation caused by incursions of barbarians headed for Rome. Instead of traveling to the earlier cathedral of St. Lorenzo, which was nearby but outside the Roman city walls, Florentines may have wanted a new cathedral within those walls.

Nonetheless, the placement of St. Reparata is still surprising when one considers the urban context of Florence fifteen hundred years ago. Roman Florentia was laid out checkerboard-style, aligned with the cardinal points of the compass. In this dense environment, the cathedral’s site was obscure and peculiar. What was fine for a private palace -- a quiet district immediately inside the north city wall -- was an inferior site for the city’s cathedral, where it was dark, visually insignificant, and vulnerable to attack.

Also, with the whole city to chose from, why stick the new church so close to the old but still viable St. Lorenzo? Surely Christians from the other quarters must have grumbled plenty that both of the city’s two big churches were in the same quarter.

Not only was the site poor, the building’s construction was somewhat odd. In design it was a standard basilical plan with an apse. The eastern half of the structure was built from the first as a church, while the western half had quite evidently been converted from the earlier palace, which originally dated to the first century A.D. The building crew had gone to exceptional lengths to preserve two and probably three walls of the old domus to provide the shell of the new church.

I had initially imagined that the builders had reused the old walls for economy, but anyone who has done both new construction and retrofitting of old buildings knows that working with old structures is actually wasteful of time and money. It seemed contradictory that the Florentines would have spent a small fortune on floor mosaics and then disfigured the exterior of their cathedral by insisting on reusing the scarred and crooked walls of the pre-existing palace. There must have been some burning reason why the new church retained a shape that deliberately reminded worshippers that it had once been a house.

Had this once been a pagan house of worship that had been taken over as a sign of Christian triumphalism? Unlikely.

Had a rich widow disposed of her mansion as a gift to the church? There were numerous earlier cases in which a wealthy widow or a Pope tore down their homes and paid to have a basilica constructed over it. But by the sixth century, the Church had rules against consecrating all but the best of these “gift” churches, and then only if they came with endowments. It is improbable that any donor would have bamboozled the bishop of Florence into accepting a compromised building on a compromised site.

We know of many cases in which a miracle, a martyrdom, or the presence of a holy person pushed an ordinary house into the realm of sacred architecture. Was it some such event that turned this house into a church? The archaeological data were encouraging to such a hypothesis: within what was probably the main bedroom of the house someone had installed a sacrario, or tiny shrine, complete with pilgrim’s oil flask. This was done before the house was turned into a church. Such shrines were uncommon in late antique houses: they typically designated a place of holiness.

This placement of a shrine does not mean that the Florentine domus immediately became a church. In Rome, for example, the palace in which St. Cecilia suffered martyrdom in the third century was turned into a church only hundreds of years later.

Today, in fact, there stands such a church-in-the-making in Pittsburgh. Reportedly, the pious housewife Delfina Cesarespada was brought back to life in 1963 with the intercession of the Blessed Nunzio Sulprizio. The Cesarespada family members were so grateful that they put a plaque to the Blessed Nunzio on the outside of their house and turned their entire first floor into a shrine where Mass is said once a month. Were this a more spiritual era, and were there an increasing and not decreasing need for churches in Pittsburgh, the Cesarespada house would undoubtedly turn itself into a full- blown church.

Poor as the early church records are in Florence, there happens to survive a document of a momentous event in a private palace of exactly the right era and the right neighborhood. This happening may have caused the placing of a sacrario in the palace below the church of St. Reparata and the eventual transformation of the house into a church.

The momentous event involved not the mythical St. Reparata but the well-documented St. Ambrose, archbishop of Milan and arguably the most important churchman of fourth- century Europe. We know from the Vita Sancti Ambrosii that Ambrose lived in Florence between March and August of 394. He resided in the palace of a senator or a nobleman of senatorial rank named Decentius, who had a wife named Pansophia and a son, Pansophius.

During Ambrose’s stay, Pansophius suddenly died on the upper floor of the palace. His mother carried the dead body down to Ambrose’s bedroom on the palace’s ground floor, where Ambrose reportedly brought the child back from death.

This story intrigued me. Did the sacrario designate the cubiculum or bedroom in which Ambrose had stayed? I tried not to be too literal, but this palace did have an upper floor (most did not, but I had excavated the stone support for the wooden access staircase).

Another passage in Ambrose’s biography made me catch my breath. It reported that Ambrose had posthumously reappeared in Decentius’s palace in 406, when he guaranteed the Florentines their miraculous victory over the Goths who were encamped on the other side of the north city wall.

This was the famous victory of 406 that was later garbled into the legends of the mythical St. Reparata.

Senator Decentius was no figment of the biographer’s invention. I have found a second ancient text -- the Relationes of the praetorian prefect Symmachus -- that mentions a nobleman Decentius just a decade before “our” Decentius played host to Ambrose in Florence. It is highly probable, though certainly not inevitable, that both references are to the same man. That would make a near-perfect match between the literary and archaeological evidence for this site. The palace I excavated below the Duomo almost has to be the one belonging to Senator Decentius. The house was rich enough and of the right time and place to fit the home of a senator, and it bore two marks of sanctity: the sacrario and its later transformation into a church.

We can at last understand why this particular palace was selected as the cathedral site, despite its disabilities. And it finally makes sense why the Florentines, when they transformed this palace into a cathedral a century later, sacrificed aesthetics to symbolism and deliberately kept its outside walls as a talisman. The barbarians had been repulsed once through the agency of those walls -- if they ever returned, the remnant of the palace would repel them again.

There are objections to my hypothesis, of course. If Ambrose was the holy person associated with the site, why wasn’t the later cathedral dedicated to him and not to Reparata? Politics may explain the snub. Perhaps Milan was jealous of sharing the fame of its archbishop with Florence; Rome surely would have been jealous of Milan’s getting credit for the victory against the Goths. After all, Florence had at least been saved in 406 from the invaders, but Rome was devastated by other Germanic tribes just four years later.

Enter Reparata. The involvement of Ambrose with Florence is beyond doubt, yet it still leaves room for, and even demands, the invention of the mythical St. Reparata. Years of visiting her two dozen other cult-sites in Italy and France has convinced me that, as in the case of so many saints, Reparata is an allegorical designation that was later personified into a fictitious woman.

This sort of personification also happened with Ste.-Foye at Conques, who started out as Holy Faith, but was later misperceived as a young martyr named Faith. The early Christian designation of Hagia Sofia (Holy Wisdom) in Istanbul was anthropomorphized in the later Middle Ages into a saint named Sophia. So also the early allegorical designations of churches to Holy Peace and Holy Victory were twisted centuries later into the fictive saints Irene and Victoria.

Reparata in Latin means “rehabilitated” or “repaired.” Her name was her game: following Florence (which remains the earliest of the miracle-sites) numerous cities acclaimed Reparata every time they were liberated from external threats, whether these threats came in the form of Goths, Saracens, Vikings, earthquakes, political strife, or the plague. There are Reparata churches in the Abruzzi and on Sardinia that were originally dedicated to Liberata, which makes the wordplay still more obvious. In Florence, too, Reparata was often called “Liperata,” which is halfway between Liberata and Reparata.

The invention of St. Reparata was possibly helped along by a Constantinian motto that circulated on coins for centuries. This was felix temporis reparatio -- “happy times are recovered.” A politician of today might render the motto as “let the good old times roll again.” Who could have resisted a saint with political muscle like that?

So a mystery that Dante’s contemporaries and modern scholars despaired of solving is now solved. And Florence, so richly endowed with the literary, scientific, and artistic genius of Dante, Giotto, Leonardo, Michelangelo, and Galileo, gets visible and tangible evidence of something it forgot it had also had: a touch of sanctity.

It may be paradoxical that the solution to the old question of the origins of Florence Cathedral came from the United States. But medieval scholarship is like any other kind of scholarship in the humanities or the sciences: it requires method and evidence, though fascination with legends and myths probably does not hurt.

Franklin Toker, a past president of the Society of Architectural Historians, is professor of the history of art and architecture at the University of Pittsburgh.

He directed the archaeological excavations below the Cathedral from 1969 to 1974 and the outside excavation in 1980. He is writing a four-volume archaeological history on the Florence Cathedral district for the Pontifical Institute of Mediaeval Studies in Toronto.

He received NEH funding in 1979 (excavations), 1985, and 1992 (research support).