Andrea della Robbia

Andrea was born in Florence on 20 October 1435, the son of Marco della Robbia. Marco's brother Luca was already the leading sculptor of the family, first trained and celebrated for work in stone and bronze before he began experimenting with glazed clay sometime in the 1440s. Andrea trained under Luca from an early age and had already worked alongside him for more than twenty years by the time he produced his first major independent commissions in the 1470s.

The workshop was also the family's home, a house on Via Guelfa in Florence about a ten-minute walk from the Duomo. When Luca's health failed in the 1470s Andrea effectively took over direction of the business before his uncle's death, and on Luca's death in 1482 he inherited half the house and workshop, eventually becoming sole owner. Where Luca had produced smaller, more restrained reliefs, Andrea pushed the technique into architecture at a scale his uncle never attempted: friezes, fountains, wash basins, facade roundels and multi-section altarpieces assembled from dozens of separately modelled, fired and glazed pieces of clay. He kept accepting major commissions into his seventies, with documented work running to at least 1515, ten years before his death in Florence on 4 August 1525.

The scale of the operation is easy to understate. A single large altarpiece could require dozens of individually molded, hollowed, fired and glazed sections, each one planned so its joins would read as part of the architecture rather than as seams. Religious houses, hospitals, confraternities and aristocratic families across Tuscany all wanted what one modern curator has called a mark of status on their buildings, and clay was cheap: what a patron paid for was the workshop's skill, not the raw material.

The technique itself is well documented by direct scientific analysis of surviving pieces. The clay body is a pale-firing, chalky material the workshop mined from a source along the banks of the Arno that later writers describe as closely guarded; unlike ordinary red terracotta it fires to a buff color, which makes the glaze laid over it read as more luminous, and its rate of expansion and contraction through firing is matched to the glaze, reducing the risk of cracking. Conservators at the Metropolitan Museum, examining two Andrea della Robbia reliefs after treatment, found large voids, uneven wedging and lumps of unmixed clay inside the fired body: evidence that the workshop's own craftsmen worked the material with far less finicky care than a modern ceramicist would, trusting an intimate, practiced knowledge of exactly how far the clay could be pushed.

The glaze itself is a tin-opacified lead glaze: tin oxide for the dense, brilliant white, with cobalt for blue, copper for green and manganese for the plum-purple that recurs across the family's work, confirmed directly by scanning electron microscopy and X-ray spectroscopy on pieces at both the Metropolitan Museum and the Fine Arts Museums of San Francisco. Large works were first modelled whole, then cut into sections designed to fire safely (twelve on the Metropolitan's Saint Michael the Archangel lunette, fifteen on its Prudence tondo), cast from plaster piece molds, hollowed to an even wall thickness to prevent warping, fired once to harden the clay and a second time after the colored glazes were brushed on. Workshop numbering systems, scratched into unglazed clay before firing to record the correct order of assembly, and even finger marks and tool impressions left in soft clay, still survive on pieces that have never been fully dismantled since installation.

None of this was decorative indulgence. Glazed terracotta cost a fraction of carved marble or bronze casting for the same surface area, its colors read clearly from across a nave or a piazza in a way unglazed stone or fading fresco could not, and, most importantly for exterior work, the fired glaze was effectively weatherproof: the Ospedale degli Innocenti roundels and the Saint Michael the Archangel lunette, made for a church facade in Faenza, both survived centuries of direct outdoor exposure with their color intact. For a religious order or hospital that wanted a monumental, legible, durable image and could not afford marble, glazed terracotta was not a compromise. It was the better material for the job.

Giorgio Vasari recorded a story that a woman working in the della Robbia household and shop leaked the family's glaze secrets to Benedetto Buglioni, who went on to run a rival glazed terracotta workshop in direct competition with the family that had, on this account, invented the technique. It is a good story and it has been repeated for four centuries. It is also, on the evidence available, probably not what happened.

Modern accounts of the two workshops note instead that Buglioni most likely trained with the della Robbia before setting up on his own business, which would make him a departing apprentice rather than a thief, and that no serious rivalry between the two firms is documented beyond ordinary market competition. A separate, more fundamental question about the technique's origin is unresolved on different terms: some scholars argue that Luca della Robbia did not invent tin-glazing from nothing but revived and adapted a technique already established among Islamic and Hispano-Moresque potters, applying it for the first time at monumental sculptural scale. Whether that argument or a genuine independent Florentine invention is correct is not settled in the sources consulted for this record, and this record does not arbitrate it.

What is documented, rather than legendary, is that the della Robbia family did guard its methods closely and successfully for more than a century. Within a relatively short span after the deaths of Andrea's sons Giovanni (1529 or 1530) and Girolamo (1566), the family's carefully protected technical knowledge of the clay source and the glaze recipe was lost, and no workshop after them ever fully reproduced the specific combination of clay, glaze and firing that had made della Robbia terracotta recognizable at a glance for a hundred years.

Andrea's most famous single work is also his most functional. In 1487 he installed ten glazed terracotta roundels, each about a meter across, on the facade of the Ospedale degli Innocenti, the foundling hospital built to Filippo Brunelleschi's design on Florence's Piazza della Santissima Annunziata. Each roundel shows a single infant in swaddling bands against a blue ground, one arm freed and reaching out, as if beckoning the viewer toward the entrance below. No two are alike: the babies differ in pose, in how far they have worked an arm loose from the wrapping, in expression, giving ten variations on the same basic design rather than ten repetitions of one mould.

The choice of subject was not decorative whimsy. A foundling hospital's entire civic and charitable purpose was announced, in a single glance from across the piazza, by an image of an infant safely bound and protected. The swaddling itself, standard Renaissance infant care rather than a symbolic invention, became through repetition on this one building a recognizable civic mark: the bambini of the Innocenti are still used today as shorthand for the institution itself, more than five centuries after they were fired. That durability was the point of the material as much as the meaning was the point of the image. White figures on a saturated blue ground are legible at a distance in a way carved stone relief rarely is, and the glaze has needed no repainting or restoration of its color since the fifteenth century, a claim few facade decorations in any medium can make.

The richest single body of documented, in-situ Andrea della Robbia commissions anywhere is not in Florence but on the remote Apennine ridge of La Verna, the Franciscan sanctuary where Francis of Assisi received the stigmata in 1224. Between about 1476 and 1493 Andrea's workshop supplied five glazed terracotta altarpieces for La Verna's three principal devotional buildings: the Annunciation (about 1476) and the Adoration of the Child (1479, dated by an inscription on its tempietto balustrade) in the Chiesa Maggiore, later joined there by the Ascension of Christ (by 1493, originally the church's own high altarpiece, moved to a side chapel in 1601); the Madonna della Cintola, installed by 1488, in the small entrance church of Santa Maria degli Angeli; and the Crucifixion, dated 1481, in the Chapel of the Stigmata itself, the most sacred space in the complex.

These were not anonymous gifts. Four of the five altarpieces carry inscriptions naming the Florentine families who paid for them: the Niccolini arms flank the Annunciation's inscription, Jacopo Brizi of Pieve Santo Stefano is named on the Adoration, the Alessandri family's arms appear on the Crucifixion and the Bartoli and Rucellai families' crests, tied to an intermarriage completed by 1488, appear on the Madonna della Cintola's predella. The Ascension alone carries no inscription, and is attributed on documentary grounds to Medici patronage of the church's presbytery and choir. Most tellingly, the Crucifixion altarpiece's date is fixed not by style but by an entry in the monastery's own account book for August 1481 that names the patron, Tommaso Alessandri, and Andrea della Robbia together: a rare instance in his whole career where a document, rather than a stylistic argument, puts his name on a specific dated work.

A 2021 study by the art historian Stephanie Miller argues that the Annunciation and Adoration altarpieces were designed with a specific Franciscan devotional experience in mind, encouraging pilgrims arriving at what a 1493 visitor's account called another Jerusalem in the West to read the two images together and to recite prayers, such as the Angelus, whose text echoes the altarpieces' own inscriptions. That is Miller's argument, citable to her and open to revision, not a fact recovered from a document; what is documented independently is the Observant Franciscan community's own emphasis, distinct from the rival Conventual Franciscans, on the Rule of poverty and on devotion to the Virgin, which the choice of an Annunciation and an Adoration as the first two altarpieces in the rebuilt church plainly serves.

Two of Andrea's works at the Metropolitan Museum have, by accident and by conservation, given up more about how his workshop actually operated than a document ever has. In the early hours of 1 July 2008, the museum's Saint Michael the Archangel lunette, made about 1475 for the facade of a church in Faenza and on display above a gallery doorway for over a decade, fell from its wall mount and shattered into fragments ranging from tiny glaze flakes to pieces weighing several pounds. The reconstruction, painstaking and lengthy, gave conservators a view into the clay's interior structure that no intact object could offer: fingerprints left when wet clay was pressed into a plaster mold, tool marks from paddling and scraping, and a repaired fifteenth-century firing flaw where a piece of drapery had separated from the body and been reattached with a fresh coat of glaze that then fused the two pieces together in the kiln. That original repair had held for roughly five hundred and forty years before the 2008 fall undid it.

A second discovery came without any accident at all. When conservators cleaned the Museum's Prudence tondo ahead of a 2016 loan, they found numbers carved into the unglazed clay beneath centuries of overpaint on the inward-facing edges of its garland sections, a coded sequence that had been ignored, and the garland's fruit consequently mis-arranged by type rather than by the workshop's own intended order, for the ninety-five years since the Museum bought it in 1921. Once decoded and consulted against colleagues in Florence who had dismantled similar in-situ della Robbia garlands, the sequence let the Museum reassemble Prudence's garland in the arrangement Andrea's workshop had actually specified, alternating fruits rather than grouping them, for the first time since the object left Italy.

Prudence is also a documented case study in exactly how attribution moves inside this corpus. The tondo, about a meter and a half across and weighing some 350 kilograms, was catalogued as a work of Luca della Robbia from its first appearance in the modern literature until the art historian John Pope-Hennessy reattributed it to Andrea in his monograph on Luca, published in 1980, arguing from stylistic details: the posture of the figure, the treatment of the drapery folds, the set of the eyelids. Nothing documentary settled the question either way; a connoisseur's argument moved the object from one man's catalogue to another's.

The same movement runs in the opposite direction elsewhere in this archive. This publication's own record for the Metropolitan Museum's Assumption of the Virgin, the Marquand Altarpiece, describes an attribution that spent most of the twentieth century at workshop level, Workshop of Andrea della Robbia, before the Museum's own curators elevated it to Andrea's personal hand for a 2026 exhibition, on a specific technical or archival basis the Museum has not published in the sources available to that record. Read together with Prudence, the pattern is the point: individual objects move between Luca, Andrea, workshop-of-Andrea and, further down the family tree, Giovanni, not as a settled hierarchy but as a live argument that different scholars have reopened in 1980, in the 1990s and again in 2026. No cataloguer working on this family, including Andrea's own family historian discussed below, has ever closed that argument for good.

The single most consequential piece of scholarship on Andrea della Robbia was written by a man who was not primarily an art historian at all, and whose interest in the family began with an object already in this archive. Allan Marquand (1853-1924) trained as a philosopher at Johns Hopkins under Charles Sanders Peirce before returning to Princeton, where in 1883 he was made the university's first professor of art history and went on to found its Department of Art and Archaeology. His father, Henry Gurdon Marquand, was a New York collector, railroad financier and, from 1889, the Metropolitan Museum of Art's second president; in 1882, while serving as the young museum's treasurer, Henry Marquand gave it an Andrea della Robbia altarpiece, the Assumption of the Virgin, bought that same year from a Florentine dealer.

By Allan Marquand's own account, that gift is what set him on the della Robbia family for the rest of his life. He published his first scholarly analysis of his father's altarpiece in the American Journal of Archaeology in 1891, and over the following decades produced what amounts to the family's first systematic catalogue: Della Robbias in America (1912), Luca della Robbia (1914), Andrea della Robbia and His Atelier (1922, in two volumes, as Princeton Monographs in Art and Archaeology XI) and, published after his death in 1924, The Brothers of Giovanni della Robbia (1928). Erwin Panofsky later cited Marquand's della Robbia volumes among the reliable body of art-historical scholarship known to European researchers before the Second World War. Marquand's own catalogue of Andrea's atelier remains, a century on, the founding reference point against which the modern standard, Giancarlo Gentilini's two-volume I Della Robbia: La Scultura Invetriata nel Rinascimento (Cantini, Florence, 1993), is measured; what specifically Gentilini's catalogue revises from Marquand's counts is not detailed in the sources consulted for this record.

Marquand died in 1924, two years after completing his catalogue of Andrea's atelier, without knowing that the very altarpiece that had started his life's work would still be having its attribution actively argued more than a century later. He is not a footnote to that argument. He is one of its authors.

Sources differ on exactly how many of Andrea's sons worked in the family trade, five by some counts and seven by the account given in the 1911 Encyclopaedia Britannica, but they agree on which two mattered most. Giovanni della Robbia, the most productive of Andrea's sons, carried the workshop's Italian production forward until his own death in 1529 or 1530, continuing to supply the same market of Tuscan churches, hospitals and private patrons his father had built. Girolamo della Robbia took a different path entirely: in 1528 he left for France and spent nearly forty years in the service of the French crown, decorating the Chateau de Madrid in the Bois de Boulogne for Francis I with terracotta medallions and friezes fired at kilns he set up at Suresnes, later working at Fontainebleau and, after Francis I's death in 1547, on the royal monuments at Saint-Denis. A relief bust of Francis I now at the Metropolitan Museum is catalogued jointly to Girolamo and to another of Andrea's sons, Luca della Robbia the Younger, evidence that more than one of Andrea's children followed the family trade abroad, not only Girolamo alone.

Girolamo died in France in 1566, the last of Andrea's sons still active in the trade. Within what the Metropolitan Museum's own conservators describe as a relatively short time after that, the family's closely guarded knowledge of its clay source and its exact glaze recipe was lost. No later workshop, in Italy or anywhere else, ever fully reconstructed the specific combination that had made della Robbia terracotta instantly recognizable across a church facade for more than a century. What survives is what was already made: a corpus scattered across Tuscan churches, hospitals and confraternities, and, from the late nineteenth century onward, across American and northern European museum collections built by a wave of Gilded Age and Progressive Era donors, Henry Marquand among the earliest of them, who prized exactly the plain, legible, durable qualities the workshop had built the technique to deliver in the first place.

Andrea della Robbia's reputation moved through distinct phases rather than rising steadily. Nineteenth-century art historians, by the Metropolitan Museum's own conservators' account, tended to treat della Robbia terracottas as mass-produced decorative objects, and early scholarship focused overwhelmingly on attribution and iconography rather than on how the workshop actually made things; only conservation science from the 1870s onward, and especially technical studies published from 2016 onward around the first North American museum exhibition dedicated to the family, reframed the della Robbia as sculptors whose backs and undersides carry as much evidence as their fronts.

The American collecting wave that Henry Marquand's 1882 gift opened continued for decades, and one Sotheby's account of the trade dates its real surge specifically to the 1920s: the Metropolitan Museum bought a Head of a Youth in 1903 with its Rogers Fund and Prudence in 1921 with the Joseph Pulitzer bequest, received further Andrea della Robbia reliefs from the estate of Adele Lehman in 1965 and from the Edith and Herbert Lehman Foundation in 1969, and in 1960 bought the Saint Michael the Archangel lunette at a Parke-Bernet auction in New York after it had passed through a Florentine noble collection, a German collector's estate and an American industrialist's collection in turn. The same appetite reached Luca della Robbia's work as well: a Madonna long kept in the parish church of Santa Fiora in the Tuscan Maremma, confiscated by the comune itself in 1866 under the nineteenth-century suppression of religious houses, passed through Belgian and Berlin dealers before the Albright Art Gallery in Buffalo bought it in 1929, and it resurfaced at Sotheby's New York in January 2021, selling for $2,016,500 against a public appeal from the comune of Santa Fiora to bring it home rather than let it leave Italy again. The appeal did not succeed.

That is the wider pattern this record sits inside: hospitals, convents and parish churches across Tuscany that could not maintain, insure or in some cases even keep their della Robbia terracottas, and a market, first American and now global, that has kept paying to take them elsewhere ever since.