From Nicola di Maestro Antonio to Carlo Crivelli and back again

by Amanda Hilliam

The only article dedicated to the Venetian painter Carlo Crivelli (c.1430/5 – c.1494) ever to appear in The Burlington Magazine was published in March 1913.[1] The attribution to Crivelli of a newly-discovered Madonna and Child (fig.1)., which had recently passed from Duveen Brothers to the Philip Lehman collection in New York, was endorsed by the Magazine’s co-editor, Roger Fry.[2]


Fry was in raptures (fig. 2). ‘This is a work of Crivelli’s prime’, he wrote. ‘Indeed, it would be hard to name another design in which he shows quite such mastery as he does here. There is hardly another work in which the sequence of lines is so suave, the flow so uninterrupted, or in which the movements of the figures harmonize so perfectly.’ And yet, not only is the work, now in the Metropolitan Museum of Art in New York, in fact by Nicola di Maestro Antonio d’Ancona (1448-c.1510/11), a contemporary of Crivelli’s also active in the Marches, but most of the original paint has been lost, and what Fry saw were the results of a heavy restoration campaign.[3] The ‘fine condition’ he described belonged to the recently re-painted image of the Virgin and Child in a sweetly Crivellesque style, which commentators up until Ronald Lightbown in 2004 tend to assign to Crivelli’s early phase.[4] The purpose of this post is not to expose Fry, or any of those commentators. Rather, Fry’s 1913 article offers a way in to exploring almost one hundred years of art-historical scholarship relating to a single painting.




The painting’s recent history begins several years prior to Fry’s article, when it was in Baron Michele Lazzaroni’s collection in Paris. It was probably in Paris that the painting received its Crivellesque makeover, since Lazzaroni – a Roman collector, dealer and restorer – had a reputation for reworking his pictures extensively, either operating on them himself or employing the famous Sienese forger and restorer, Federico Icilio Joni, to do so on his behalf.[5] It must have been quite a job. For the Metropolitan Museum’s 1987 catalogue of Lehman’s collection, the painting was examined under the microscope and it emerged that figures of the Virgin and Child are modern (save the Virgin’s left hand, the Child’s right leg and a few other small areas).[6] The gilding and pastiglia are largely modern, and only some details in the microarchitecture of the throne are original. Freshly re-painted and ready for sale, the work then moved to Duveen Brothers in New York, perhaps on the initiative of Berenson, who was an agent for Joseph Duveen, procuring Italian paintings primarily from Lazzaroni, and by late 1912 it was in the collection of Philip Lehman. [7]

Duveen was evidently eager for the newly-discovered Renaissance masterpiece, which had recently passed through his hands, to gain the approval of the scholarly community. On 21 November 1912, Duveen wrote to Berenson announcing the sale of the painting to “Mr. Philip Lehman here in New York, who is a new collector and who apparently has a great ambition to add to his collection the finest pictures he can buy.”[8] But Berenson reserved his praise for the painting until 1915,[9] and it was instead Frank Jewett Mather who first published it in Art in America in the January of 1913. Mather writes that the ‘splendid Crivelli, of which Mr. Philip Lehman has become the fortunate possessor, has so recently been brought to my attention that I cannot pretend to have studied it exhaustively.’[10] The attribution to Crivelli, then, cannot have come from Mather. It must have come either from Duveen, perhaps via Berenson, or even from Lazzaroni.

One reason for Duveen encouraging a scholarly response to the painting was to dispel any doubts surrounding its authorship, of which there must have been several, for, unusually for Crivelli, this once-central panel of a polyptych is unsigned, and the gothic shape of the parapet and the receding checkered tiles are not found in any other painting by Crivelli. Fry addresses this point directly: ‘The work is not signed,’ he writes, ‘but of its authorship no doubt appears possible. It has in a supreme degree the delicacy of Crivelli’s contour as well as the firmness and brilliance of his painting.’[11]

Although the authentication was accepted and had a lasting impact, the close parallels with Nicola di Maestro Antonio’s work were noted almost immediately, first by Berenson in 1915,[12] and then in 1927 by Franz Drey, author of the second monograph on Crivelli, who remarked that the curved parapet was typical of Nicola di Maestro Antonio but the only example in Crivelli’s oeuvre.[13] In 1958, Federico Zeri also observed parallels between Nicola’s sacra conversazione now in the Palazzo Massimo alle Colonne, Rome (1475-1480) (fig. 1) and the Lehman Madonna, even if, according to the author, the latter had been ‘revitalised with the improvements of a famous collector-dealer in Paris fifty years ago,’ thus confirming Lazzaroni’s intervention. [14]

Unperturbed by these hints, however, the authors of the later monographs on Crivelli (Anna Bovero, 1961; Pietro Zampetti, 1986; Ronald Lightbown, 2004) continued to give the work to Crivelli without qualms.



For Bovero, the painting predated Crivelli’s first signed work, the Massa Fermana polyptych (1468), and was a sure sign of his interactions with Antonio and Bartolomeo Vivarini in Venice.[15] While for Zampetti, the painting was to be dated several years later, to between 1470 and 1473, for its closeness to the Madonna for the 1472 polyptych for Fermo (fig. 3).[16]

Unlike Bovero and Zampetti, who note Berenson’s and Zeri’s opinions, Lightbown does not mention the painter from Ancona in his discussion of the painting, instead opting for something that compares to Fry’s rambling formal analysis: ‘Once more we find the suave fusion of breathing tenderness with grace and majesty so characteristic of Crivelli’s Madonnas as this splendid moment of his art, when it reaches its first apogee of dramatic intensity of invention and potent mastery of expressive form.’[17]

This was despite John Pope-Hennessy’s warnings in his entry in the 1987 catalogue of Lehman’s collection, when the painting’s ‘deplorable’ condition was finally taken into account. Pope-Hennessy offered two explanations. The first, suggested to him verbally by Zeri, ‘of regarding the present Madonna […] as a ruined work by Nicola di Maestro Antonio, which was repainted before 1913 to resemble the work of Crivelli in the 1470s.’ And the second, ‘of considering it a ruined early work by Crivelli, painted under the influence of Schiavone, not of the Vivarini, which was repainted in the style of Crivelli’s mature paintings. The first of these explanations is likely to be correct.’[18]

Indeed, it was.

In 2008, ninety-five years after the painting was first published, Matteo Mazzalupi proposed a firm attribution to Nicola di Maestro Antonio on the basis of compositional parallels with known works by the painter and traces of the quinquefoil moulding from the painting’s original frame.[19] Mazzalupi lists four other panels by Nicola which were framed in identical quinquefoil mouldings in Avignon, Baltimore and Jesi, suggesting tentatively that the panels once formed part of a polyptych commissioned in 1481 for the convent of San Francesco in Staffolo.[20] The painting’s entry on the museum’s website reflects this revised attribution, which was first put forward by Zeri, and advanced by Pope-Hennessy and Mazzalupi.

What can we learn from this episode, if anything? The prolonged incorrect attribution to Crivelli of the Metropolitan Museum’s panel is symptomatic of the fact that art historians (traditionally, at least) are often not as conversant as they should be on issues of condition, materials and techniques. Panofsky’s error in reading Lucas Cranach the Elder’s Cupid removing his blindfold as based on Plato, on account of the volume inscribed Platonis opera upon which Cupid then stood – which, however, was removed during cleaning in 1973, revealing the original stone plinth underneath – is but one example of the perils of failing to consider how a work of art may have changed over time.[21]

Access to conservation technology and the expertise of conservators and scientists certainly helps, but above all the art historian needs to understand how to read the physical signs left on the object they are studying, like a barb from an unusually shaped frame, before making art-historical judgements on iconography, attribution or otherwise. It is of course easy to look back at the errors of those before us with the privilege of hindsight, as future art historians will undoubtedly do to us. But what we can do is interrogate stylistic evidence, rather than blindly accept it, and remember that the works of art themselves are our most important documents.


Fig. 1. Nicola di Maestro Antonio, Madonna and Child Enthroned, c.1487?, Metropolitan Museum of Art, New York

Fig. 2. Fry’s article in the Burlington Magazine

Fig. 3. Carlo Crivelli, Madonna and Child Enthroned, 1472, Metropolitan Museum of Art, New York



[1] Roger Fry, ‘Madonna and Child by Carlo Crivelli’, The Burlington Magazine for Connoisseurs, vol. 22, no. 120 (1913), pp.308-309.

[2] Roger Fry and Lionel Cust were joint editors from December 1909 until December 1913, when they were joined by More Adey. See, Caroline Elam, ‘A More and More Important Work: Roger Fry and The Burlington Magazine’, Burlington Magazine vol. 145, no. 1587 (2003), pp. 142-152.

[1] Roger Fry, ‘Madonna and Child by Carlo Crivelli’, The Burlington Magazine for Connoisseurs, vol. 22, no. 120 (1913), pp.308-309.

[3] For the work in the Metropolitan Museum, see (accessed February 2017).

[4] Ronald Lightbown, Carlo Crivelli (New Haven and London: Yale University Press, 2004), p.203. For the complete literature on the painting, see Matteo Mazzalupi, ‘Nicola di maestro Antonio’ in Pittori ad Ancona nel Quattrocento, eds. Andrea De Marchi and Matteo Mazzalupi (Milan: F. Motta, 2008), cat. 14, p.291.

[5] Colin Simpson, Artful Partners: Bernard Berenson and Joseph Duveen (New York: Macmillan, 1986), note 18, p.89.

[6] John Pope-Hennessy, The Robert Lehman Collection, I, Italian Paintings (New York: Metropolitan Museum of Art in association with Princeton University Press, 1987), p. 228.

[7] Edward Fowles, Memories of the Duveen Brothers (London: Times Books, 1976), pp.122-123.

[8] Cited in Pope-Hennessy, p.228.

[9] Bernard Berenson, ‘Venetian Paintings in the United States: Part Two’, Art in America, April 1915, Vol. 3, No. 3, p.113.

[10] Frank Jewett Mather, Jr, ‘A Madonna by Carlo Crivelli’, Art in America Vol. 1, January 1913, pp. 48-53.

[11] Fry, p.308.

[12] Bernard Berenson, ‘Nicola di Maestro Antonio d’Ancona’, Rassegna d’arte antica e moderna, vol. 15 (1915), pp.168, 171, 173-174.

[13] As cited by Mazzalupi, pp. 257, 289.

[14] Federico Zeri, ‘Qualcosa su Nicola di Maestro Antonio,’ Paragone, vol. 9, no. 107 (1958), p.40.

[15] Anna Bovero, Tutta la pittura del Crivelli (Milan: Rizzoli, 1961), p.54.

[16] Pietro Zampetti, Carlo Crivelli (Florence: Nadrini, 1986), p.257.

[17] Lightbown, p. 203.

[18] Pope-Hennessy, p. 228.

[19] Mazzalupi, pp. 257-259.

[20] Ibid., 259.

[21] Paul Taylor, Condition: The Ageing of Art (London: Paul Holberton Publishing, 2015), pp.9-10.




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s