Carlo Saraceni, also known as Carlo Veneziano, was born in Venice and arrived in Rome about 1598. In the years leading up to his election to the Accademia di San Luca in 1607, Saraceni primarily produced small cabinet paintings. One of the first he executed in Rome is The Met’s oil on copper Paradise
), which, despite its modest scale, took inspiration from a large altarpiece by Francesco Bassano and ceiling fresco by Niccolò Cirignani, both in the Jesuit Church of the Gesù in Rome. His style developed in response to key artists then working in the city, notably Adam Elsheimer, a German resident in Rome between 1600 and 1610, and Caravaggio. Between 1616 and 1617, under the patronage of Pope Paul V, Saraceni collaborated with Giovanni Lanfranco and Agostino Tassi in the Sala Regia of the Palazzo del Quirinale (see 68.51
). In 1620, the last year of Saraceni’s life, he returned to Venice. The Patron:
Laerzio Cherubini (1556?–1626) was a successful lawyer and jurisconsult who also acted as a civic magistrate and legal historian. His self-made fortune was built from real estate holdings and his role in drawing up loans on revenues from the Roman curia. Devout and civically minded, Cherubini served on the governing board of the Casa Pia, a charitable institution for disadvantaged women that owned the site on which the Church of Santa Maria della Scala was built in Rome. Four of Cherubini’s six sons became monks and another became a canon in the church of Santa Maria in Via Lata.The Picture’s Genesis and Commission:
The Met's picture is indelibly linked to the painting it replaced, Caravaggio’s canvas of the same subject (Musée du Louvre, Paris; see fig. 1 above). On June 14, 1601, Caravaggio signed a contract with Laerzio Cherubini to paint “the death, or transit, of the blessed Virgin Mary” (mortem sive transitum Beatae Mariae Verginis) to measurements that would fit Cherubini’s recently completed chapel in Santa Maria della Scala. The church, built between 1592 and 1610, belonged to the Discalced Carmelites, an order whose independence Pope Gregory XVIII had established only in 1580. “Discalced” comes from Latin, “without shoes,” indicating their vow of service to the poor; the Carmelites’ charism, or spiritual focus, was contemplation and prayer, as it remains today .
There is little reason to suspect that Caravaggio failed to meet the conditions of his contract, which stipulated completion of the painting within one year, by June 1602. Pamela Askew notes that Cherubini’s will, dated August 4, 1602, implies this chronology. It specifies that in his recently completed tomb for the chapel in Santa Maria della Scala, designed by architect Girolamo Rainaldo, the phrase “Sacred to the Virgin” (Deiparae Virgini sacrum) be inscribed “above the painting” (super quadrum), suggesting that Caravaggio’s painting was already in place.
Letters from the art collector Guilio Mancini to his brother Deifebo indicate that by October 1607 the Carmelites were looking to sell the painting, ridding themselves of it on the grounds of its “inappropriateness in lasciviousness and decorum” (spropositata di lascivia e di decoro) and their conclusion that it was “without decorum, invention, or cleanliness” (senza decoro, e invention e pulitezza). Some years later, Mancini would hint at Caravaggio’s possible use of a prostitute as a model for the Virgin. Despite Mancini’s efforts to broker the purchase of Caravaggio’s painting, by the winter of 1607, Peter Paul Rubens, working in Rome under the employ of the Duke Vincenzo I Gonzaga, recommended it in the highest terms for acquisition by the Mantuan court. The duke’s ambassador to Rome, Giovanni Magni, negotiated the purchase. Before it was shipped to Mantua on April 28 (in a crate Rubens had constructed), Magni put the painting on public view for the week of April 1–7, evidently in order to satisfy members of the Accademia di San Luca. During this time it prompted public scrutiny and, on the whole, praise. No documentation of direct exchanges between Caravaggio and Cherubini or the Discalced Carmelites is known today, but the objections to the painting were almost certainly lodged not by Cherubini but by the order itself. The order housing the chapel would have had ultimate authority over issues of iconography and doctrine proper to the site. Indeed, Magni’s correspondence indicates that it had “been refused by the church to which it was donated” (rifiutato dalla chiesa a quale era stato donato). As Patrizia Cavazzini has uncovered, painters held little leverage in these situations. Approval of such large-scale commissions and even the final price paid by the patron were typically determined after a work was completed. Giovanni Baglione and Giovan Pietro Bellori, contemporaries who would have had access to the facts even if they wrote years later, followed Mancini in terms of the Carmelites’ rationale in rejecting Caravaggio’s Dormition
. They blamed above all the Virgin’s corporality: Caravaggio had “with little decorum made the Madonna swollen and with bare legs” (con poco decoro la Madonna gonfia, e con gambe scoperte) and she appeared “too much like the swollen corpse of an ordinary, dead woman” (troppo imitato una Donna morta gonfia).
The fortune of Saraceni’s replacement for Caravaggio’s rejected work helps clarify how Caravaggio had failed in the eyes of the Discalced Carmelites. That Saraceni, too, was developing a naturalism not far from Caravaggio’s suggests that this kind of pervasive stylistic shift in Italian painting in the early seventeenth century was not what had caused offense–or at least not primarily. Rather, issues of doctrine and iconography around the subject of “the death, or transit” of the Virgin stated by the initial contract were most likely at the core of the controversy. By the time of Saraceni’s commission, the situation would have been reinforced in 1603 by Cardinal Camillo Borghese’s revival of an edict asserting the authority of the Church to judge the suitability of the paintings found in its chapels.
While no contract for the second commission is known to survive, Saraceni could have been approached to replace Caravaggio’s painting as early as 1607. Our earliest document regarding his involvement is from Fioravante Martinelli’s (1599–1667) manuscript “Rome ornamented by architecture, painting, sculpture” (“Roma ornate dall’architectura, pittura, escoltura”) generally dated between 1660 and 1663, some fifty years after the event. He records of Santa Maria della Scala: “the second chapel on the left of Laertio Cherubino [Laerzio Cherubini], compiler of the Ballario Romano [legal documents published in 1617] and celebrated criminal lawyer [who] had Carlo Venetiano [as Saraceni was also known] make a beautiful painting with the transit of the Madonna with Apostles and other figures and background of an [architectural] perspective. The work was in place for little time and greatly praised, but because the Fathers wanted a Glory of Angels to replace the perspective background, he took the picture back; he sent it to Venice and in a few days produced another for which he was paid 300 scudi” (la seconda capella à mano manca di Laertio Cherubino compilatore del Ballario Romano, et celebre avvocato criminale haveva Carlo Venetiano fatto un bellissio quadro col transito della Madonna con Apostoli, e altre figure, e di sopra con una prospettiva. Fu posto in opera per poco tempo e grandemente lodato, ma peché i Padri vi volevano la gloria con angeli in cambio di detta prospettiva, egli si ripigliò il quadro; che fu mandato a Venetia, et in pochi giorni fece quello che hora si vede pagato col prezzo di scudi 300).
Martinelli’s account elucidates aspects of the surviving works: there is The Met's version, which clearly underwent modification by Saraceni early on, as well as a second painting, with a Glory of Angels, that remains in situ as part of the decoration of the Cherubini Chapel (figs. 2–3). Visual and technical analysis of The Met's painting reveal that the Virgin’s head originally slumped forward, her eyes were closed, and her hands were folded on her bosom rather than upright in prayer. In short, hers was a resolutely dead body. Presumably, she appeared much as she does today in the reduced replicas Saraceni and his workshop subsequently produced in oil on copper (see, for example, fig. 4). It is tempting to surmise that Martinelli’s recollection of the first painting whisked off to Venice and replaced by a new canvas in merely “a few days” was mistaken and does not describe the final work hanging in Santa Maria della Scala today. Given the labor required to produce the new painting, replete with the raised head, open eyes, and Glory of Angels instead of architecture, a more likely scenario is that he refers here to Saraceni’s attempt to rework The Met version. Regardless, his 300 scudi (Caravaggio had received only 280 scudi, but the higher amount may reflect this additional labor and even materials for a second canvas) was given upon production of the painting still in situ. The austere, classical architecture of the first canvas (in its original and modified states) had presumably been intended to reference the concept of the Virgin’s body as the body of the Church. This interior, not so different than Santa Maria della Scala itself, was perhaps deemed too contemporary, however, and insufficiently otherworldly by the Carmelites who were concerned with the Virgin’s body appearing too mortal and not obviously en route to its heavenly incarnation. Not only did the interior architecture’s replacement with a Glory of Angels make the Virgin’s transit to heaven clear, but the angel crowning her with flowers further alludes to her mystical marriage to Christ upon arrival in heaven. As Christiansen has noted, a more subtle change captures this shift between documenting an earthly event and the translation of that event into theological discourse: in addition to the Apostle Peter, who is lost in thought in the foreground, a mourning figure at left in the first version was replaced in the final version by an apostle who seems to discuss the significance of what occurs before him.
This series of changes indicates that the wording of Caravaggio’s contract from 1601, “the death, or transit, of the blessed Virgin Mary,” likely remained for Saraceni, its ambiguity still producing iconographic confusion. What the Carmelites actually required, yet oddly seem not to have specified from the start for either artist, was not a representation of the Virgin’s death in any earthly, corporeal sense, but rather a miraculous transition of the Virgin through death to heaven: death as a transition, The Transit of the Virgin
, or as the painting is known today, The Dormition of the Virgin
. The rejection of Caravaggio’s painting as well as Saraceni’s successive changes support the iconography as the primary dilemma for the Carmelites, rather than issues of style, including the naturalism so associated with Caravaggio and his followers. In fact, the distinctive features of the bald man kneeling at left would have been recognizable among the Caravaggisti as a favorite model, a Slavic man whom Guido Reni was said to have discovered along the Tiber river and admired for resembling a famous statue thought to represent the philosopher Seneca. His features, including pointed ears and nose, are well documented in paintings by Reni, Bartolomeo Manfredi, Orazio Borgianni, and particularly Jusepe Ribera, making it possible to recognize his subtle presence here.
Saraceni produced at least four small-scale, oil on copper versions of The Dormition of the Virgin
, found today in the Alte Pinakothek, Munich (inv. 185), the Gallerie dell’Accademia, Venice (inv. 679), and two private collections (fig. 4). Apart from variations to the architecture, color of the costumes, and framing of the scene, they all are closest to The Met's painting in its first state: they all lack the Glory of Angels and include the Virgin’s head slumped, her eyes closed, and hands folded on her chest. In a practical sense, the architectural background (“the perspective”) was surely less labor intensive for paintings produced for the art market. The fact that these repetitions retain features of the first version of the painting, however, further supports the hypothesis that the Discalced Carmelites’ objections were specific to their own order and not offensive in a broader context. Michele Nicolaci notes that these repetitions of Saraceni’s first version may also indicate the painting was in place longer than the “little time” mentioned by Mancini, and for some Romans it was his “official version” of the painting. Use of the first version of the composition in later repetitions may also be due to the workshop’s reliance on Saraceni’s initial series of drawings and oil sketches.
N. Randolph Parks proposed that the figure staring out of the far left of The Met's painting was a donor portrait depicting Cherubini. The absence of this figure from the copper reductions–where it is replaced by a figure with hands clasped overhead in anguish–reinforces this supposition. Less explicable, however, is the fact that in Saraceni’s final version for Santa Maria della Scala, this figure looks out of the painting, in the manner of a donor portrait, but appears far younger (in 1612 Cherubini would have been about fifty-six years old, an age more appropriately represented in The Met's painting). Based on Saraceni’s Self-Portrait
(Accademia di San Luca, Rome), Nicolaci has suggested that the figure may be a representation of the artist himself.
Martinelli’s manuscript states that “A print was made of the first [version of the painting]” (Il primo và volta stampato in rame). Indeed, the etching to which he must refer, made in 1619 by Jean Le Clerc (41.97.461
), follows neither the Santa Maria della Scala nor The Met's version as it exists today, but the copper reductions with the Virgin’s body resolutely dead. As Annette Hojer observed, Le Clerc’s print comes closest to the copper version now in Munich, dated 1619 and mentioned in Saraceni’s will of 1620 as having been a commission for Sebastian von Füll auf Windach on behalf of Maximillian I, Elector of Bavaria. This work, already a reduction, was surely easier to translate into print and was easily accessible in Rome, where Le Clerc was active.
At some point in its history, approximately eighty centimeters appear to have been removed from The Met's canvas, as indicated by the difference in size of the work in situ and the absence of the uppermost vault, which is visible in the backgrounds of the reduced versions in oil on copper.Commissioning Conflict in Caravaggio’s Rome:
In an attempt to counteract the spread of Protestantism in Europe, at the Council of Trent (1545–63), the Catholic Church issued stricter rules for the production of paintings. Article 25 of the Council stated: “let great care and diligence be used herein by bishops, as that there be nothing seen that is disorderly, or that is unbecomingly or confusedly arranged, nothing that is profane, nothing indecorous, seeing that holiness becometh the house of God.” Cardinal Camillo Borghese’s edict of 1603 asserting the authority of the Church to judge the suitability of the paintings found in its chapels was a reverberation of another portion of the same article, which stipulated that “no one be allowed to place, or cause to be placed, any unusual image, in any place, or church, howsoever exempted, except images that have been approved of by the bishop.”
As Caravaggio and Saraceni’s attempts to satisfy the commission of Laerzio Cherubini for Santa Maria della Scala attest, added institutional pressure and shifting ground regularly resulted in conflict between artists, their patrons, and the orders whose churches would house a given painting. The frequent rejection of works, perhaps due to the orders’ insecurities or uncertainties about dogma, also resulted in the circulation of site-specific works, notably large-scale altarpieces, on the international European art market from early in these paintings’ histories. While art historians might be eager to locate this conflict in issues of style–Caravaggio’s naturalism, his daringly theatrical composition and lighting, for example–in most cases an artist surely was chosen in part because their way of painting was appreciated. By 1601, a well-informed Roman should have had a sense of the visual qualities of Caravaggio and, indeed, the sensational value and force with which they might propagate Catholicism. More concrete problems of iconography were, instead, the most likely driving force behind the rejection of works of art such as Caravaggio's Dormition
, the precise subject of which the Carmelites, or Cherubini, seem to have inadequately conveyed. The many oil on copper reductions of the initial, rejected version of Saraceni and his workshop’s Dormition
and the circulation of this image in print attest to the fact that this composition was not blasphemous in a universal sense, but merely understood as improper by the Discalculated Carmelites who controlled Santa Maria della Scala.
The chronology proposed above places the completion of Caravaggio’s Dormition
in the midst of some of his most pivotal and controversial early works that in turn set the stage for the obstacles faced by Saraceni. The peculiar nature of Caravaggio’s rejections and the immediate afterlife of his paintings reveals a changing sense of the role of works of art and, indeed, of artists, connoisseurs, and collectors. The first version of Caravaggio’s Crucifixion of Saint Peter
(lost), executed in 1601 for the Cerasi Chapel in Santa Maria del Popolo, Rome, was rejected but was found two years later in the collection of the Marchese Giacomo Sanesio. Caravaggio’s Inspiration of Saint Matthew
(destroyed in World War II), itself commissioned to replace failed sculptures by Jacques Cobaert for the Contarelli family chapel in San Luigi dei Francesi, Rome, was rejected in 1602 and acquired by the famed connoisseur Vincenzo Giustiniani. As Christiansen notes, one of the most remarkable things about these rejected commissions is the fact that Caravaggio’s paintings actually accrued value on the art market almost immediately and that they were purchased by leading collectors such as Sanesio and Giustiniani. Similar phenomena surround the moment in which the Carmelites put Caravaggio’s Dormition
on the art market in 1607. The artist’s Madonna and Child with Saint Anne
(1605–6; Galleria Borghese, Rome), commissioned as an altarpiece for the Archconfraternity of the Papal Grooms in the Basilica of Saint Peter’s, Rome, appears to have caused sufficient alarm that it was soon exhibited instead in the Vatican parish church of Sant'Anna dei Palafrenieri. Ultimately, however, it landed in the collection of Cardinale Scipione Borghese for a higher price than the initial commission. By the time Caravaggio had fled to Naples in 1606, his newly completed Madonna of the Rosary
(Kunsthistoriches Museum, Vienna), evidently commissioned by a donor (whose portrait is included at left) but rejected from a Dominican order, was being considered for purchase by the Duke of Mantua. While he would ultimately decline, he would, of course, purchase Caravaggio’s Dormition
on Rubens’s suggestion within the year at a price that fully reimbursed Cherubini for the 280 scudi that he had paid to Caravaggio for his failed canvas, the highest fee of Caravaggio’s Roman career. Notably, as per Caravaggio’s contract, the money paid to the artist was set neither by Cherubini nor the Carmelites but by the connoisseur and collector Vincenzo Giustiniani upon inspection of the finished work which, in spite of the iconographic issues it raised, he evidently deemed an outstanding work of art.
David Pullins 2019
 For further details on Cherubini’s life and affiliations, see especially Askew 1990.
 Parks 1985; Askew 1990, pp. 3–18.
 Askew 1990, pp. 6–7.
 Michele Maccherini, "Caravaggio nel carteggio familiare di Giulio Mancini," Prospettiva
, no. 86 (1997), pp. 76–78.
 On his use and this trope, see Frances Gage, "Caravaggio’s 'Death of the Virgin,' Giulio Mancini, and the Madonna Blasphemed," in Caravaggio: Reflections and Refractions
, New York, 2017.
 Walter F. Friedländer, Caravaggio Studies
, New York, 1969, pp. 307–10.
 Friedländer 1969, p. 308.
 Patrizia Cavazzini, "'Patto fermo' o cortesia negli accordi tra pittori e committenti a Roma nel Seicento," Ricerche di storia dell’arte
, no. 101 (2011), pp. 5–20.
 Giovanni Baglione, Le vite de' pittori, scultori et architetti: dal Pontificato di Gregorio XIII del 1572 in fino a' tempi di Papa Urbano Ottavo nel 1642
, Rome, 1642, p. 138.
 Giovanni Pietro Bellori, Le vite de' pittori, scultori et architetti moderni
, Rome, 1672, p. 213.
 Christiansen 1992, p. 300.
 Marciari 2010, pp. 140–45; Nicolaci 2013, pp. 210–14. Nicolaci speculates that, at this stage, Saraceni would have been required to submit a bozzetto
, or sketch, of some kind for approval of a complex iconographic scheme, particularly after the failure of Caravaggio’s painting.
 The manuscript is generally dated between 1660 and 1663. Martinelli 1660–63 in D’Onofrio 1969, p. 134; Askew 1990, p. 65.
 Christiansen 1992, p. 301; Nicolaci 2013, p. 214.
 Christiansen 1992, p. 301.
 Viviana Farina, Al sole e all'ombra di Ribera: Questioni di pittura e disegno a Napoli nella prima metà del Seicento,
vol. 1, Castellammare di Stabia, 2014, pp. 78–82; Keith Christiansen, Valentin de Boulogne: Beyond Caravaggio,
exh. cat., The Metropolitan Museum of Art, New York, 2016, pp. 53, 78, 85.
 For another version, perhaps with workshop assistance, see Dorotheum 2015. For a partial discussion, see Nicolaci 2013, pp. 216–20; Hojer 2013 pp. 318–20.
 Nicolaci 2013, p. 212.
 Parks 1985, p. 443.
 Inv. 577. See Nicolaci 2013, p. 215; Terzaghi 2013, pp. 344–45 in same exh. cat.
 Martinelli 1660–63 in D’Onofrio 1969, p. 134.
 Le Clerc’s etching seems to have been pirated in an anonymous print, also from the seventeenth century (62.602.414
). As part of larger campaigns to publish the Bavarian State Collections, the Munich copper was further reproduced in etching by Joseph Fischer (62.602.413
) in the early nineteenth century, and as a lithograph by Johann Nepomuk Strixner in the 1810s (British Museum, 1854,1020.1362), and by Johann Wölfle in the 1860s (British Museum, 1866,1013.500).
 Hojer 2013, p. 318.
 Nicolaci 2013, p. 214.
 Christiansen 1992, p. 298.