Printer Friendly

Antonello's lost 'St Augustine' rediscovered: Joanne Wright reveals a major discovery: an altarpiece panel by Antonello da Messina, depicting St Augustine. Now in a private collection, it was formerly known only from a photograph owned by Bernard Berenson.

Amongst Berenson's 'homeless paintings' published in the Final edition of Italian Pictures of the Renaissance (1957) was a panel ascribed to Antonello da Messina and described as a 'full-length Augustine. E. Lfor early]', with the illustration captioned as 'signed' (Fig. 4). (1) No provenance is given, and no indication of how long it had been unlocated when Berenson recorded it. Neither is there any information about the origins of the photograph from which he worked, which dates from the early 20th century and is still at I Watti. (2) On its reverse, however, there are several inscriptions, one of which (in Mary Berenson's hand) records that the work had been for sale in Rome in 1909. (3) Although he did not publish it until 1957, Berenson must have had the photograph in his possession before 1945, as Mary died in that year. He may have acquired it from an art dealer who would have been able to supply the information about its appearance on the Roman art market.


In 1980 Giovanni Previtali summarised the then current view of Berenson's attribution when he wrote that it had not been widely accepted, but that Berenson had not had the opportunity to defend his opinion, save to say that he considered the work to be early. (4) Previtali clearly did not agree with this consensus, for he referred to St Augustine as 'splendid' and pointed to an inscription in relief Roman capitals along the shaft of the bishop's crozier (5) that Berenson had noted and which, as an unusual (but not unique) form of Antonello's signature, Previtali felt suggested authenticity. (6) A pasticheur, he argued, would have been inclined to use the much more common form of a hand-written signature on a cartellino, a carefully-unfolded piece of paper apparently attached to the front of the picture plane, a device that Antonello very much made his own. And, Previtali continued, to propose that it might be the deliberately ambiguous signature of Antonello's nephew Antonio de Saliba would be to ignore the difference in quality between the work of the two artists, even when De Saliba was at his best. In his deliberations over attribution Berenson had also considered De Saliba, but rejected him in favour of his more gifted uncle.

The panel of St Augustine, known for over half a century from this old photograph alone, has now come to light in a private collection in the USA (Fig. 2). (8) It has been cleaned, and Previtali's view on the De Saliba issue is dearly right. Even De Saliba's greatest achievement, the Enthroned Madonna (Castello Ursino, Catania, Fig. 3), painted in 1497, although certainly competent, is slightly pedantic and formulaic, demonstrating the great debt he owed to his uncle rather than presenting a qualitative challenge to his work. St Augustine, by comparison, has a freshness and directness lacking in De Saliba's work.


Even working from a photograph, Previtali was perspicacious enough to note that the background and the halo seemed to have been repainted. He reserved judgement on the fabric and decorated border of the cope. (9) Cleaning has proved Previtali right. The halo and background were indeed painted over a beautiful landscape, rendered with both the fluidity and loving attention to detail of Antonello's Flemish mentors, coupled with a characteristic sense of volume and space. The cleaning of the figure and the drapery has also enhanced our reading of the work. The modelling, which in Berenson's photograph looks rather stodgy, is transformed in many areas into a crisp vitality. Much of the detail is revealed as exquisitely painted, with that highly-focused attention to the quality and nature of surface minutiae associated with Flemish painting of the period, here grafted onto a more robustly modelled Italianate form.

The saint's face (Fig. 1) has been considerably damaged, but sympathetic retouchings made in the most recent conservation have softened its expression. Any remaining sense of the features being slightly mask-like is the result of the extensive in-filling that was necessary in certain areas. Worst affected in this way are the foremost cheek just above the beard and again below it on the lips, the far cheek below the eye, and most significantly, the area around the eyes themselves. (10) But generally the new re-touchings are more subtly blended into the surrounding flesh tones than those removed in the conservation process, and much of the over-defined linearity visible on Berenson's photograph has vanished. Otherwise, the figure is in a generally good condition. What is most remarkable is the high quality of finish that remains largely intact, since in so many of Antonello's works extensive surface abrasion has occurred. The lively treatment of the beard and eyebrows contrasts tellingly, for example, with the vestigial bland fuzz on the chin of the much-damaged St Benedict from the S Gregorio Polytych (Museo Regionale, Messina, Fig. 5) and is comparable with the 'whiskery' eyebrows of the Trivulzio portrait (Museo Civico, Turin).


Previtali's caution with regard to the status and condition of the drapery, which, he maintained, could be authenticated only by direct examination, was well founded. Again, cleaning has revealed significant changes to the image presented in Berenson's photograph. The dark surplice now tapers to a soft triangular fold at its bottom edge, improving the compositional flow of the lower part of the panel. The white cassock below it now lies in softer, subtler folds, the starkly linear contrast between white and black in its former modelling becoming a more gently modulated composition of greys and pearly whites. What looked more like crumpled paper before cleaning, has now taken on the verisimilitude of softly falling fabric. Although on a different scale, it bears comparison with the falling drapery of the angels' cassocks in the Three Angels fragment, Reggio Calabria (Fig. 9) or the Magdalen's drapery in the Crucifixion recently moved from Bucharest to Subiu (Fig. 10). As to the cope's rich brocade, the most startling difference as a result of cleaning is the revelation of a beautifully-figured, blue-grey damask lining whose subtle modulations of tone capture the fall of light and shade across it (Fig. 12). Its outer surface is of a smaller and rather more tightly-conceived pattern than many of Antonello's brocade draperies. But it is handled in a similar way and bears a remarkable resemblance in the pattern of the neat daisy-like flower to the lower expanse of brocade worn by the Madonna in the S Cassiano Altarpiece fragment (Kunsthistorisches Museum, Vienna, Fig. 11). The mitre, which is particularly richly be-jewelled, is otherwise conceived and executed in a way comparable to those of St Gregory and St Benedict from the S Gregorio Polyptych (Museo Regionale, Messina) and those of the three half-length Doctors of the Church (Galleria Regionale, Palermo). In all of these, the combination of precision in the depiction of the gems and the looser, more flowing brushstrokes that define the pattern in the brocade, is striking.


As is so often the case, it is in the handling of detail that we fred the artist's 'signature'. The understanding of the structure of the head and facial features, the responsive vitality of the engaging, light-reflecting eyes and the physical formation of the slightly sensuous mouth are all characteristic of Antonello, found extensively in his work, but lacking in that of his followers. The relationship of the back of the neck to the fabric of the collar lining the cope (Fig. 6) is directly comparable with the same passage in the St Benedict from the S Gregorio Polyptych, as is St Augustine's hand holding the volume of his writings (Fig. 7). The way in which the sleeve of the cassock bisects the back of the hand just below the thumb, forming a long triangle of white between the flesh and the dark, heavy fabric of the cope, is also used to very similar effect in both figures. The same is true of the articulation of the fingers of the other hand around the crozier. A pentimento is visible where the thumb of the saint's right hand has been changed (Fig. 8). Originally it stood out from the crozier, bent back in a characteristic Antonellian gesture such as that, for example, of St Benedict in a panel acquired several years ago by the Castello Sforzesco, Milan (Fig. 15), or that of the S Gregorio Polyptych St Benedict. On the surface, however, St Augustine's thumb is wrapped round the crozier to form a dosed fist holding the shaft, akin to the hand of St Nicholas in the S Cassiano fragment. (11)


Berenson stated that the panel was 'early'. However, Mary Berenson appears to have inferred the date 1473 from its proximity in composition to St Benedict from the S Gregorio Polyptych (see note 3). Previtali also draws on this comparison, but uses it to provide a terminus ante quem of 1473. He then argues that the 'Flemishness' of the drapery is similar to that in the Sibiu Crucifixion and the Three Angels fragment and proposes to date it after these two but before the S Gregorio Polyptych--between 1465 and 1470. (12)

Microscopic and ultra-violet examination of the painted inscription on the crozier could not distinguish it from the paint in the surrounding area, and this further seems to put the seal on its authenticity (or at least its contemporaneity with the rest of the work). And what neither Berenson

nor Previtali could see from the photograph is that the inscription is actually completed with a date (Fig. 13). The full inscription runs down the lower part of the crook and is only partially legible. (13) It can be reconstructed as OP [US] ANTONELLUS (SIC) MESS[ANEUS] [?]4[?]0.[14) The date's first numeral, which is totally obliterated, obviously must be a '1', but the third is both marginally more legible and much more intriguing. What at first glance appears to be a '7' is, on closer inspection, revealed to be damage (Fig. 14). (15) What then might this third numeral be? All that can be discerned is the start of a downward sloping bar placed precisely equidistantly between the '4' and the '0' and possibly the start of a curve half-way down and to the tight of the bar (Fig. 14). By a process of elimination it must be either a '5' or a '6'. (16)


A date of 1450 is improbably early, as it would make St Augustine one of the artist's earliest extant works, as early as, if not earlier than, the Sibiu Crucifixion, which in both style and handling is less sophisticated and assured than this panel. (17) A date of 1460, however, bears closer examination. It has already been noted that the panel demonstrates many of the stylistic and genetic characteristics of Antonello, although the slight tightness of handling has been used to call its authenticity into question. But I believe that this might better be associated with the restraint of an artist not yet at the height of his confidence when working on this scale. (18) At 1460, the panel would fit into what both Berenson and Previtali identified as Antonello's early, most Flemish-influenced phase. The drapery falls in a similar fashion, although on a grander scale, to that in the two Reggio panels of the Three Angels and St Jerome in Penitence, both apparently from the late 1450s. The minutely-observed hairs of the beard and eyebrows, the rendering of the gems on the mitre and the cope and the intricacies of the patterns both on the outer surface of the cope and on its lining--all these show particular attention to lessons learnt from contemporary Flemish art. The painting of detail in the landscape foreground is also characteristic of Antonello's early style, seen again in the Reggio panels. And the general disposition of the landscape, with St Augustine standing on a foreground hummock and the mid-ground falling away below to form a continuity with the distance, follows the Eyckian formula also used in the Sibiu Crucifixion. (19)

The surface quality of the paint appears somewhat less refined than in works from Antonello's later career. Gas chromatographic testing of several samples from St Augustine reveals that the binding medium is linseed oil. Antonello was well aware of, and experienced in, the use of oil by 1460, although the evidence of early works suggests that he seems to use a pigment-tomedium ratio that leaves the paint texture fairly thick and sticky. Later in his career he appears to have learnt that paint was easier to manipulate if it was more dilute, and seems to have added more oil to the pigment, making it more fluid and at the same time more capable of fluency in handling. It may be that the 'clotted' consistency of his paint is responsible for a little awkwardness in handling, visible in passages of the Sibiu Crucifixion and the Three Angels fragment. And, on a larger scale, the slightly unrefined quality in St Augustine (discussed below) could at least partially be a result of this same early-career pigment/medium ratio. (20)

Compositionally, the panel's closest comparison is the figure of St Benedict from the S Gregorio Polyptych, with which Mary Berenson clearly wanted to associate it when she suggested a date of 1473. But St Benedict demonstrates much greater fluidity and the assurance of a mature artist fully confident of his ability to work on the polyptych scale. (21) St Augustine is stiffer, less at ease with his pose, more self-conscious--the very qualities that a youngish artist rising to the challenge of representing a standing saint on a larger scale than that to which he was accustomed might well display. They are also the qualities one might expect to be resolved as the artist grew in maturity over a period of 13 years. In other words, it does not seem unreasonable to suggest that the artist who painted St Augustine in 1460 should mature into the artist who painted St Benedict in 1473.

The compositional similarities between St Augustine and St Benedict from the S Gregorio Polyptych are such that one is a virtual mirror image of the other. Rationalisation of the drapery to reveal the feet and thereby define rather than obscure the stance, a slight inclination of the head of St Benedict and an alteration to the angle of the crozier as it lies across the drapery are the only significant changes. The two figures are similar in scale, although not identical, and the ratio between the various parts is extremely close. This suggests that a single drawing was utilised as a template for both figures, scaled-up appropriately for each. Antonello's involvement in altarpiece painting over the years required him to provide a number of variants on the figure of a standing ecclesiastic. The surviving examples (St Benedict and St Gregory from the S Gregorio Polyptych, the St Benedict from Castello Sforzesco, Milan, and St Dominic and St Nicholas of Bari from the S Cassiano Altarpiece) show that he used and re-used a limited number of variations. Why should he not reverse and revise an earlier version in a later altarpiece? It would have offered an opportunity to 'improve' on his earlier attempt, relaxing the figure painting and giving it added assurance.

Previtali argued that a panel of a tonsured saint usually identified as St Bernard, Abbot early part of his career. The few extant attributions, and those works lost but recorded in documents, are almost exclusively panels for processional banners (gonfalom) or small icons and devotional panels. Amongst these I would include the much damaged, but sensitively-handled Sta Rosalia (Walter Art Gallery, Baltimore) but exclude The Salting Madonna (National Gallery, London), a work of lapidary hardness and odd proportions which is, in my opinion not by Antonello at all--a view apparently shared by James Beck (review of the of Clairvaux (Fig. 16) in a Florentine private collection must belong to the same complex as St Augustine because of the similarity of dimension and subject. (23) This is, in my view, mistaken, for the apparent similarities between the panels are, in fact, very superficial. The younger saint inhabits a different world from that of St Augustine, on whom he is indubitably modelled. His head is much less animated and his proportions are conceived quite differently, being shorter in the arms and longer in the legs. And whereas St Augustine's cloak articulates his anatomy, showing the position of the elbow and the precise recession of the right arm, the anatomy of the tonsured saint is shrouded in drapery which focuses on complex folding rhythms without offering any rational explanation for its fall. (24)


An inscription, which seems to be based on a knowledge of that on St Augustine's staff, apparently reads opus ANTONELLENIUS 1490, a date that undermines Previtali's argument that this panel and St Augustine belong to the same project. (25) The certainty that the third numeral on St Augustine's staff cannot be a '9' must separate these two works as surely as their stylistic and qualitative differences do. (26) Furthermore, why should two panels ostensibly from the same project both be signed? The Florentine panel is surely from Antonello's studio, possibly by Salvo d'Antonio, and is modelled on the type established by the older artist a generation earlier. (27) Its artist pays homage to his dead master not only in the use of his standing bishop saint type, but in the imitation of a conceit devised by him years earlier, the signing of the work on the staff. The workshop did little but pay tribute to Antonello in the decades following his death, churning out 'Antonello style' altarpieces and polyptychs for churches particularly in the northern and eastern regions of the island.

The conclusion must be that St Augustine is by Antonello. Dated to 1460, it provides vital information about the character of the artist's work at the end of his 'early' period, a phase bedevilled by documented but lost work, dubious attributions and the poor condition of a number of potentially key paintings. As its surface, although damaged, is in much better condition than, for example, St Jerome in Penitence (Reggio Calabria) or Sta Rosalia (Baltimore), it can be seen to be an excellent example of his early work, representing the climax of his first full decade as an artist. It also seems to be his first extant altarpiece painting, for it surely once formed part of a composite altarpiece. The composition and provenance of this polyptych will be discussed in the second part of this study, to be published in APOLLO in March.

This article (in both its parts) was first presented as a paper at a London National Gallery. Renaissance Seminar. I am indebted to Lorne Campbell, Jill Dunkerton and Luke Syson, whose participation in the discussion afterwards helped to improve and refine parts of the argument. I am also grateful to my colleague Jeremy Wood, who read and discussed an early draft with me, and to Professor Francis Ames-Lewis, who asked many probing questions.

(1) Bernard Berenson, italian Pictures of the Renaissance, Venetian School, Cleveland and New York, 1957, vol. I.p. 7 and plate 277.

(2) Fig. 4 is reproduced by kind permission of Dr Fiorella Superbi, Fototeca Berenson, Villa I Tatti, Florence, who confirmed its approximate date.

(3) The first inscription, in Berenson's hand and in pencil, is partially erased and simply reads 'Antonello'. Over this erasure, written in pen by Mary, 'Antonello?' signifies a later doubt about the attribution. This is amplified by a further inscription in Mary's hand: 'if he, 1473 but note ear and thumb of left hand. For sale in Rome 1909. Ear of Berlin and Davis' Saliha lip idem and Winthrop's. Thumb of Vir. An. at Venice'. Below this, in red pencil, Berenson later added, 'Antonello. Signed on staff Opus Anto Mess'.

(4) Giovanni Previtali, 'Da Antonello da Messina a Jacopo di Antonello, 1. La Data del 'Cristo Benedicente' delia National Gallery di Londra', Prospettiva, no. 20, 1980, pp. 27-34.

(5) There are two other painted inscriptions on the panel in an 18th-century hand that apparently refer to the Messinese artist Girolamo Alibrandi (1470-1524). Alibrandi can have had nothing to do with this panel, which is patently quattrocento in character. An obfuscation of the picture's origins may have been attempted in the late 18th century, perhaps after the 1783 Messina earthquake, when looting was commonplace.

(6) There are similar illusionistically-incised inscriptions attesting authorship on the front ledge of both the Malaspina portrait (Museo Civico, Pavia) and the portrait of a young man in the Pondazione Thyssen--Bornemisza, Madrid, where it is, however, truncated horizontally.

(7) The cartellino signature does not, however, appear in Antonello's work until at least 1465, if one accepts that date for the Christ Blessing (National Gallery, London), or even later if one does not.

(8) It was acquired by its present owners in 1972 as 'una tavola rappresentante un santo'.

(9) Previtali, op. cit., p. 34, note 15.

(10) The creases around his fight eye, although slightly heavy-handed (surely the result of re touching), resemble in character the lines on the fight of the brow of the S Casssiano St Nicholas. Their prominence here mars the surface; the face is enlivened when this area is covered.

(11) A paint loss on the second-version thumb reveals the painted staff lying beneath it, indicating that this change was made after the painting of that area was substantially finished. Like the pentimento in Antonello's Christ Blessing (National Gallery, London), it reminds us that the artist's working method was not so rigid as to preclude late changes of mind, and that he went through a phase where he was unaware of the increasing translucency of oil glazes with the passage of time.

(12) Previtali, op. cit., p. 30. I have argued for a much earlier date (early 1450s) for the Sibiu Crucifixion and, in the same article, suggest a date of the later 1450s for the Three Angels fragment: J. Wright, 'Antonello in formazione: un riesame della Crocifissione di Bucharest', Arte Veneta, no. 45, 1996, pp. 20-31.

(13) Antonello's signing methods were unconventional by Italian standards of the day. In the first place, he attached his name to his work much more commonly than most of ins contemporaries did. Then, inspired by his Flemish mentors, he inscribes parapets beneath portraits or, more frequently, uses the trompe Voeil cartellino. He even adopts this very casual format for ins S Gregorio Polyptych. None of Antonello's other extant altarpiece panels bears an inscription. This may be because the), are fragmentary (e.g. the Viennese S Cassiano fragments) or because large areas of the paint surface have been erased (e.g. the Syracuse Annunciation) or because the inscription was on a different panel (now lost) from the same altarpiece (e.g. the Correr Pieta). it might also be that, like the majority of altarpieces of tins period, they were not signed at all.

(14) The fact that the Latin is grammatically incorrect perhaps further supports the contention that the inscription is original. Antonello's grasp of Latin seems to have been less than secure, judging by his inability to equate the date and indiction of the London National Gallery's Christ Blessing. Anyone charged with adding this inscription later would presumably have been given a grammatically correct text.

(15) This is dearly visible using a hand-held magnifying glass.

(16) It certainly excludes the possibility of an '8' or a '9', which any argument involving Antonio de Saliba (born c. 1466) or Salvo d'Antonio (born after 1461) would demand. T. Pugliatti, Pittura del Cinquecento in Sicilia. La Sicilia Orientale, Naples, 1993, p. 52, attributes this work to a collaboration between these two nephews of Antonello, c. 1490.

(17) Wright, op. cit., pp. 20-31. The Sibiu Crucifixion has always been accepted as an early work, although the date has vacillated between 1450 and 1460 in attempts to associate it with Antonello's Neapolitan experience (F. Bologna, Napoli e le rottle mediterranee della pittura. Da Alfonso 1, Il Magnanimo a Ferdinando il Cattolico, Naples, 1977) or to connect it with a documented gonfalone commission (M. Rostworowski, 'Trois tableaux d'Antonello da Messina', Jaarboek, Koninklijk Museum voor Schone Kunsten, Antwerp, 1964). Longhi suggested the idea (unlikely for such a small work) that it was worked on in two distinct phases ('Frammento Siciliano', Paragone, vol. 47,1953).

(18) Antonello (b. c. 1430) seems to have focussed on smaller-scale works in the exhibition 'Antonello da Messina', APOLLO, June 2006, p. 89) but not by the curators of the recent exhibition in Rome. A small panel (15 x 10.7 cm) recently acquired by the Museo Regionale di Messina, The Virgin and Child with a Franciscan Donor with an Ecce Homo on the verso may be his very earliest work, possibly dating from the late 1440s (see J. Wright, Christies Sale Catalogue, 9 July 2003, lot 81, pp. 149-54).

(19) This is the formula identified in M. Meiss, 'Highlands in the Lowlands: Jan van Eyck, the Master of Flemalle and the Franco-Itallan Tradition', Gazette des Beaux Arts, vol. LVII, 1961, pp. 298-301.

(20) I am grateful to Jill Dunkerton and her colleagues at the National Gallery, London for advice on the medium and its handling and helpful interpretation of the gas chromatographic samples.

(21) The St Augustine panel measures 110.5 x 39.5 cm; those of St Benedict and St Gregory. from the S Gregorio Polyptych are both 125 x 63 cm.

(22) St Augustine is precisely 100 cm from the tip of the mitre to the foot of the drapery. St Benedict measures 109.5 cm from the tip of the mitre to the feet.

(23) Giovanni Previtali, 'Alcune opere di Salvo d'Antonio da ritrovare', Prospettiva, no. 36,1983-84, p. 132 and ill. p. 125. The panel is of almost identical dimensions to the St Augustine. But many altarpiece panels of standing saints by Antonello and his followers are of this fairly standard formula and size (c. 110 x 40 cm).

(24) Further important discrepancies can be observed. The crozier (which appears to be the same studio prop in both panels) is held perpendicular by the tonsured saint, and the dynamic relationship between the shaft and the saint's body created by the diagonal in the St Augustine panel is lost. Similarly the head of the crozier relates much less well to the head of the saint, partly because of the relative proportions and placing, but largely because, by comparison with that of St Augustine, it is conceived in space in a much less articulate way. St Augustine's crozier casts a subtle shadow that extends across the drapery after crossing the ground in front of the saint. No such subtlety is evident in the Florentine panel A report by the conservator Lisa Venerosi Pesciolini (June 2003) indicates that the Florentine panel has recently been conserved; it can therefore be assumed that we are seeing it at its best.

(25) Previtali, op. cit. in n. 23 above, p. 132.

(26) It also contradicts Previtali's own view that St Augustine is, as Berenson suggested, an autograph early Antonello. Previtali, op. cit. in n. 4 above, p. 30.

(27) The facial similarity to Sta Lucia formerly at Castellmare di Stabia (and associated with Salvo since the time of Cavalcaselle's drawing of the work in the 1860s) is remarkable.

Joanne Wright formerly taught art history and directed the Djanogly Art Gallery at the University of Nottingham, where she is now Director for Enterprise in the Arts.
COPYRIGHT 2007 Apollo Magazine Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Wright, Joanne
Geographic Code:4EUIT
Date:Jan 1, 2007
Previous Article:Lesser Gods: Pontormo's fluid mind and engaging humour are revealed in two newly discovered drawings for decorative schemes commissioned by the...
Next Article:Double acts: retrospectives for the chapman brothers and Gilbert & George pose the question of how two people can be a single artist.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters