A History of the Restoration

A History of the Restoration

Ambrogio Lorenzetti, Il Buongoverno; detail with the integration by Pietro di Giovanni Orioli, 1492. Siena, Palazzo Pubblico.

A History of the Restoration

Some notes on its ancient origins
The great architect-restorer, Eugene Viollet-le-Duc, referring to the Latin terms “reficere”, “instaurare”, “renovare”, immediately specifies that these terms do not mean to restore, but to recover or make afresh. He observes, with a conscious self-satisfaction in his own set of values so characteristic of the nineteenth century, that both the concept and the practice [of restoration] are modern. The conclusions of the great architect-restorer are not belied by Pliny’s accounts of the various events and misadventures surrounding the conservation of famous works of art: when we examine the sources, they would seem to indicate that works of art were considered more as trifles (“ludicrae”) to delight the ear and the eye {“ad volup- tatem aurum atque oculorum”), as Seneca observed, rather than being instances of a figurative discourse, bearing a cultural message.

Simone Martini, Guidoriccio da Fogliano at the siege of Montemassi, detail; integrated at the end of the fourteenth century. Siena, Palazzo Pubblico.

Simone Martini, Guidoriccio da Fogliano at the siege of Montemassi, detail; integrated at the end of the fourteenth century. Siena, Palazzo Pubblico.

The problems associated with images pertaining to Christian worship are of a different nature: and this is without taking into account problems more purely anthropological, such as considerations on the nature of objects of worship and their transmission. When Origen in his Thirteenth Homily on Genesis compares man to an image painted by God, on which man himself has then painted the earthly image with its vices, like colours hiding the original paint, it becomes clear that images were preserved with appropriate repainting from the very earliest times. The practice of maintaining icons through repainting, respectful of the iconography of the image, shows us the most usual method of conserving an image; the importance of the example expounded by Origen, is that it shows that in the first half of the third century, this practice was sufficiently widespread for him to be able to extract from it an easily accessible moral fable.
Of even greater significance is what is implicit in Origen’s comparison between the original image and the one repainted by man with all his sins. With reference to the transmission of images, it allows one to clarify a concept which is even now central to the vision of restoration and its aims, and which is as deeply rooted in the Biblical tradition of the garden of Eden as it is in the myth of the Golden Age: the return to a primitive state which is better than the present one. The deep roots of this vision in both mythology and Western religious tradition allow us to understand (and this we can observe on a daily basis) how dangerous this vision can become when restoration is approached without an adequately critical spirit: indeed, it induces one to pass over the concept of the ageing of materials, and often will impose the model of a return to the original (ripristino), whatever the cost.
Moreover, transferral, survival and new context for a work of art are not all one and the same thing: we need only look at the long history of the reuse of antique fragments in the Middle Ages. Without again running through a subject that has already been adequately covered, I should only like to mention the head of Livia, which was used in the ninth century for the Herimankreuz in Cologne, because it lends itself easily to comparisons with similar salvage operations that occurred within the compass of sixteenth-century collecting, such as the head of Tiberius in the Museo degli Argenti in Florence. In its new context, the head of the Herimankreuz takes on the meaning of the head of the Redeemer: its use as 1, 2 such may have been suggested by the recognition of its formal perfection, or else by its suggestive qualities of the past glories of Rome, and the collapse of paganism and its replacement by the new Christian faith. It exemplifies how a different context can give new iconographic meaning to the recovered fragment, and imbue it with new ideological values.
Tiberius’ head, adapted in 1581 to fit a new rich and ornate gold mounting by Antonio da Faenza, may have been interpreted and treated more or less respectfully, but has nevertheless been presented according to the subject attributed to it; it testifies to the existence of an ancient world that can serve as an example, but is no longer retrievable. Restoration can only repropose iconographic or formal values, to a fragment which is in itself an artistic or a historical rarity.

Piero della Francesca, Saint Jerome; after the removal of the repainting dating from the early years of the sixteenth century. Berlin Dahlem, Staatliche Museen.

Piero della Francesca, Saint Jerome; after the removal of the repainting dating from the early years of the sixteenth century. Berlin Dahlem, Staatliche Museen.

Piero della Francesca, Saint Jerome; with the repainting dating from the early years of the sixteenth century. Berlin Dahlem, Staatliche Museen.

Piero della Francesca, Saint Jerome; with the repainting dating from the early years of the sixteenth century. Berlin Dahlem, Staatliche Museen.

The reworking (rifacimenti) of altarpieces from the end of the Middle Ages presents us with a wide spectrum of adaptations, renewals and repaintings; an artist charged with the maintenance of a painted panel would find it difficult to refrain from some little touch of repaint, perhaps to brighten the colours that the cleaning had not sufficiently revived, or else to bring the painting up to date iconographically or in line with the prevailing taste of the day. With the onset of the use of X-radiography as an analytical tool, more and more images have been discovered beneath the ones we see; a number of scholars now fear to manoeuvre their way round this minefield, after so many thirteenth-century panels have been found to have been repainted at a date insidiously close to that of their creation, leading to many errors in chronology.
For example, Coppo di Marcovaldo’s Enthroned Madonna in Orvieto serves as a lynch- 3 pin in the reconstruction of the Master’s reuvre, and it is on this work that the appraisal of the artist as a precursor to Cimabue now rests. During the recent restoration of the work at the Istituto Centrale del Restauro, it was discovered that the principal heads had been completely repainted, probably as the result of damage caused by fire. It seems to me that no other conclusion can be put forward (and in this I differ from those directing the restoration) but that the repainting was carried out in the last years of the thirteenth century by a painter who had seen the work of Cimabue. His presence can be detected especially in the neck of the Madonna, with its closed, harmonious outline resembling that of a Greek vase, and in the Romanesque, almost succulent, foliage of the crown.
The most venerated panels might be subjected to multiple repaintings, as can be seen in the half-length figure of Saint Dominic in the Fogg Art Museum (Cambridge, Massachusetts), 4 a fragment from a Sienese work painted not long after the canonization of the saint in 1233. The earliest repainting of the head dates from the decade after 1260, whilst the third (which constitutes the present image) was painted about twenty years after that, and can be attributed to the workshop of Guido da Siena. The hands have also been painted over at least once, and even the tunic has at some point been brought up to date; finally, the gilded halo seems to be a punched decoration dating from the fourteenth century.
In recent times, this type of intervention has led to various misunderstandings and difficulties in dating, as happened, for instance, after a series of errors surrounding the figure of Agostino Veracini, whose work it is impossible to recognize with any certainty in a series of reconstructions in the Greek manner. As to what the philosophy should be regarding any possible intervention on these repaintings (which should at all events be one that avoids their destruction), has been expounded on with great clarity by Giovanni Romano. It should be borne in mind that copies and falsifications of thirteenth-century paintings of a satisfactory standard are a very recent twentieth-century phenomenon (for instance, the Volpi Madonna, exhibited in 1937 in the Giotto Exhibition). The Madonna dell’Impruneta (painted in 1758 by Ignazio Hugford in imitation of the no longer visible image) is a good example of the limitations of eighteenth-century artists struggling with the style of the Early Masters, despite having a passionate devotion to their art.

Giovanni Bellini, Madonna Frizzoni; with the sky repainted in the sixteenth century. Venice, Museo Correr.

Giovanni Bellini, Madonna Frizzoni; with the sky repainted in the sixteenth century. Venice, Museo Correr.

Giovanni Boccati, Pala dei Disciplinati; detail with the repainting by Giannicola di Paolo, 1519. Perugia, Pinacoteca Vannucci.

Giovanni Boccati, Pala dei Disciplinati; detail with the repainting by Giannicola di Paolo, 1519. Perugia, Pinacoteca Vannucci.

With the onset of the fourteenth century, it became possible to link works of art on which can be seen ancient restorations with documentary accounts that give a relatively detailed picture of the work undertaken, especially in the instance of the Tuscan cities that have been more intensively studied by art historians and archivists. In Pisa, for example, we see this in the frequent references to the repairs carried out on the frescos in the Camposanto; from the earliest references in 1371 to the point when, in 1523, il Solazzino works on the Inferno, a restoration to which Vasari also refers. Restorations can be seen in Siena on works that still exist and are often of great renown, such as those carried out by artists of the stature of Duccio or Simone Martini. The latter, in 1321, repainted eight of the most important heads in the Maesta of 1315 in the Sala del Mappamondo: a complex case, not least in the chronology of its execution, on which the recent restoration under the direction of Alessandro Bagnoli should shed some light. Whether or not the repainting is linked to conservation problems, it is nevertheless an intervention that brings the work into line with more modern taste.
The most considerable reconstructions that can be considered satisfactory from the point of view of uniting the old intonaco with the new can be seen in such famous works as 7 Ambrogio Lorenzetti’s Buongoverno and the Guidoriccio da Fogliano by Simone Martini. It is somewhat surprising that between the time of Cavalcaselle and 1955, the reconstruction in Lorenzetti’s fresco was never taken into account, and that it is only during the most recent restoration that the extensive earlier intervention on Simone Martini’s masterpiece was noted. The reconstructed part of the Buongoverno lies in the section where the wall representing the well-governed city meets the wall decorated with the allegorical figures: whether it was as the result of the violence of one of the turmoils of 1356 or 1368, or the obliteration of figures linked to the old regime once the new government was in place, a significantly large area was reconstructed by Andrea Vanni, with technically excellent results.
The other fresco is, instead, an example of a complete reconstruction, including the application of a new intonaco, of the whole area representing the castle of Montemassi. It was executed in a manner so faithful to the original that no scholar, viewing it from below, had ever noticed either the difference, or the hesitations in the handling, which are now evident even in reproduction. It is likely that the offending intonaco was knocked down, once the castle painted on it had either been traced or copied in some other manner. It also seems clear that whoever executed the reconstruction still shared the Trecento vision of space: there was no correction of the characteristic perspective construction of that time, nor any overturning of the planar construction, resulting in a representation of Montemassi that is entirely in keeping with the architecture of the stories of Beato Agostino Novello.

Beato Angelico, Saint Matthew, with Lorenzo di Credi’s repainting, 1501. Chantilly, Musée Condé.

Beato Angelico, Saint Matthew, with Lorenzo di Credi’s repainting, 1501. Chantilly, Musée Condé.

Thus, this is not an iconographic reworking, as is the case in Lorenzetti’s work, but a true integration, once the damaged intonaco had been replaced, and which everything suggests is faithful to the original.
The documents in the Opera del Duomo in Siena bear witness to the continuing upkeep of the polychrome sculptures and painted panels in the cathedral. The note of a payment made to one Martino di Bartolomeo on 3 November 1404, for the panel of the altar of the wood-carvers and stone-masons: “to which I put back some colours and saints”
(“a la quale rimessi cierti colori e santi”), gives us a good idea of the spirit in which this work was undertaken. Martino’s intervention was seen as bringing the altarpiece in line with the new demands of worship, changing or adding figures of saints and recovering the colours that no longer carried out their optical function, to use the terminology of modern restoration. This intervention occurred only, and because of, the function of the panel as an object of worship, and as part of the furniture that brought prestige to the cathedral.
In Florence, the Strozzi papers (“spoglistrozziani”) record the ongoing restoration of the mosaics in the Baptistry, common maintenance practice for such decoration, whilst in 1392, Benedetto degli Albizi asked Niccolo di Pietro Gerini to “complete and repair” (“compiere e racconciare”) a Deposition in San Pier Maggiore, of which he left us an interesting record when adding “and it was painted by Maso the painter, a great master”.
Gothic polyptychs, and frames “in the Antique style”
The attendant parts complementing the works of the Early Masters, in addition to alterations dictated by changes in usage, resulted in a kind of intervention which, by the middle of the fifteenth century in Florence, was not unusual; that is the squaring up of cusped polyptychs within rectangular frames. The spandrels between sections were filled, and pilasters in Renaissance style were added, as well as friezes inspired by the new architectural style. Contemporary documents term these new frames adornments “all’antica” (in the Antique style), and it was Offner who had already drawn attention to them as examples of “early modernizations”.

Coppo di Marcovaldo, Madonna; with repainted heads dating from the end of the thirteenth century. Orvieto, Museo del Duomo.

Coppo di Marcovaldo, Madonna; with repainted heads dating from the end of the thirteenth century. Orvieto, Museo del Duomo.

Giotto’s Baroncelli Coronation was modernized in Ghirlandaio’s workshop with the 8 specific intent of preserving the figurative elements of a master who enjoyed great prestige among the humanists, and who was still the object of study by artists, as we can see from Michelangelo’s youthful drawings which can be dated from these very same years. What was no longer of any great interest was the Gothic frame, which we can imagine being similar to that of Giotto’s polyptych in the Pinacoteca in Bologna. Thepredella was preserved in 9 its original length, the main sections brought closer together, eliminating the pinnacles that divided them, and the cusped arches cut down. Only in the central section was an original part of the painting cut: the upper portion of the throne and the coronation with Leterno fra angeli, and this is the fragment that is now in the museum in San Diego, California. The sections were brought together in a beautifully carved and gilded frame, with two pilasters and a frieze decorated with heads of cherubs. More heads of cherubs were painted in the spaces created between the cusped arches and the new frame, without any loss in the quality (which one might have expected in such small-scale intervention), which is worthy of Ghirlandaio himself. The actual painting did not undergo any modernization, and the recent restoration did not find any repainting that might be ascribed to Ghirlandaio’s intervention.

A History of the Restoration

Antonio da Faenza, remounting of a head of Tiberius; 1581. Florence, Mrneo degli Argenti.

The care and intelligence with which this painting by Giotto was conserved and (in its own way) enhanced, by eliminating the Gothic woodwork, demonstrates an approach in which there is a distinction between the painting and the object as a whole; a distinction that we still have difficulty in resisting when making the decisions that are often inevitable in the field of restoration. However, it also demonstrates the high quality of these “early modernizations”, which are usually presented with a more craft-like concreteness, and preserve the devotional or heraldic-devotional character [of the work]. This is what we find in Neri di Bicci’s Ricordi which, in 1472, for instance, referring to a panel by Tommaso Soderini in San Frediano, tell us: “altered (fe’racconciare) the cusps of the arches, repainted (rifece di nuovo) four new cherubs, retouched and repainted almost all of the old figures, and turned San Frediano into Saint Margaret”.
The last great example of a polyptych adapted to fit a new frame in the style of the Antique, and by an artist highly acclaimed in his profession, is the reframing in 1501 by 10-12 Lorenzo di Credi of the panel by Beato Angelico in San Domenico in Fiesole. The sections of the old polyptych were regrouped to make a single panel, with an architectural background opening out onto a landscape which replaced the original gold, and the whole then reframed in a large frame in the Antique style. The only areas to be repainted were the throne of the Virgin, the new background and some minor adjustments in terms of perspective. The small figures of the saints, originally painted inside the old Gothic columns, were inserted within the pilasters of the new frame, adapting them to small niches in perspective. I suspect that it is only at this point in time that the Gloria Celeste (now in the National Gallery in London) became the predella of the work.
In Bologna, Francesco del Cossa repainted (rifa) the fourteenth-century fresco of the Madonna del Baraccano; he preserved the central section, and then developed an intermediary style between his own and that of the original work, which can be seen in the handling of the folds of the drapery of the central figure. However, the architecture and the landscape background are in complete accord with his painting style in that period (the fresco dates from 1472), for instance in the Griffoni Altarpiece. A similar approach can be seen in Graffione’s only documented work, the Madonna, which he adds to the Saints by Baldovinetti in Sant’Ambrogio, in Florence in 1485. This is not much help from the point of view of a stylistic identification of the artist, because of the way in which he tries to integrate his own style with that of Baldovinetti; the iconography and the compositional traits are no longer those of the earlier master, but the chiaroscuro is completely coherent with that of the original figures.

Early twentieth century fake; Madonna and Child. Formerly Florence, Volpi Collection.

Early twentieth century fake; Madonna and Child. Formerly Florence, Volpi Collection.

Giannicola di Paolo, in Perugia in 1519, found that he had partly to repaint the Pala 13 dei Disciplinati by Giovanni Boccati which he proceeded to do, constructing figures in a style not his own, making concessions to the more archaic style of the fifteenth-century master. However, in his desire to appear archaic, his handling became so awkward that it is easy to discern the vast areas of repainting.
Not so many years later, artists of different stature would soon be demonstrating quite different levels of attention [to the detail], and interpretation of styles not their own, although outside the compass of attempts at conservation or renovation of the paintings of the Early Masters. Moretto gave a perfect imitation of the Gothic style in the embroidered figures on the copes of the Doctors of the Church in the Frankfurt Altarpiece, whilst in the Saint Luke in the Church of San Giovanni Evangelista in Brescia, he inserted an icon which reproduced with complete accuracy the stylistic elements of the Greek manner. Rosso Fiorentino represented a small Madonna, in the old style, in the background of his Portrait of a Young Man in Naples, with an overall effect which is characteristic, although the poor state of conservation of the painting means that it is not possible to verify the extent to which the new consciousness of “manner”, of style, had led him to analyse and then construct the image in a stylistically different manner.
There is no doubt, however, that what we now have before us is a consciousness of “maniera” as a system to be used in the construction of an image. It would, before long, be present in Vasari, with all its implications of organic unity and homogeneity as prerequisites for any intervention that alters or makes good (risarcisca) a painting. This unity and homogeneity must be present in order for restoration to take on a different physiognomy to that of the painter’s normal activity.

Giotto, polyptych. Bologna, Pinacoteca Nazionale.

Giotto, polyptych. Bologna, Pinacoteca Nazionale.

Adaptations and renovations at the turn of the sixteenth century
Palatine Manuscript 1001 is one of the finest collections of recipes in the Biblioteca Nazionale in Florence; it is written both in Latin and in a Venetian dialect that gives us an indication of its origin, and contains various recipes directed towards the cleaning of paintings. The date of 1561 scribbled on the flyleaf allows us to turn to this text to verify which were the materials and the procedures used in the sixteenth century, and probably also in the preceding century, when a panel appeared excessively dark and one wished to “lighten” it. For example: To brighten up old gold and pictures.
“Take the finest lime and cover it with three fingers of water in a glazed container, and let it dissolve thoroughly with the aid of a stick, and then let the mixture clear. When it is clear, distil it twice with a retort and keep it to one side.
Then take ashes from soft wood and mix them together, and if the mixture is strong, so much the better. Take two inghistere [measures] of this mixture and five pounds of white soap, finely grated, and with a sponge you will make it dissolve thoroughly in this lye, and then distil it all two or three times, the more the better. Having done this, you mix together three pounds of water made up with the ashes and soap, and one of lime water; when you want to use it, first of all clean all the dust from the panel, then you will wet a sponge with the aforementioned and previously prepared tallow. Keep washing your work until it is beautiful, then with clear and pure water wash away the mixture; allow it to dry and, using cotton wool, with the remnants in the recipient, shine it up. Note, however, that having to wash the figures so that they appear clean, you must do this with such skill and in such a manner as not to remove the colour; when the painting is dry, pass over it a hand of egg-white which has not been beaten; similarly, if the egg is very thin, the mixture will be sufficient [in itself].

Beato Angelico, Saint Dominic Altarpiece; reframed by Lorenzo di Credi in 1501. Fiesole, San Domenico.

Beato Angelico, Saint Dominic Altarpiece; reframed by Lorenzo di Credi in 1501. Fiesole, San Domenico.

It is a question of being able to emulsify sufficiently well the alkaline substance with the fatty or proteinaceous components, which allow the process to be slowed down or regulated: the care and skill of the painter were of course the determining factors for keeping the operation within the desired limits. Other recipes also indicate the use of honey, presumably as a substance that would keep what would later be termed the “corrosive substances” in suspension, and limit the depth of their action. Present-day restorers use similar methods when they wish to control reagents, in order to obtain the level of cleaning that seems to them to be the most appropriate for the painting entrusted to their care:
Method of renovating paintings on panel or on walls or on gold, which are old and they will appear new:iii “Take a new, glazed cooking pot and put into it a pound of black soap and an inghistera [measure] of a mixture made of strong ash and quick lime as you know how, then boil it until the soap has dissolved, and then remove it from the fire. Have to hand a glass of strong white vinegar, three whole eggs and an ounce of common salt, and having put everything into a bowl mix it just as though you were making a broth. Take a lira’s worth of white honey, and mix it well with the other ingredients; when you are ready, you will be able, either with the finger or the brush, to quickly spread the mixture on the figures and the gold. Have the sponge to hand and wash it away, but see that the sponge is imbibed with weak lye, which is better than water. When you are wetting [the surface], make sure the mixture or the water is clean and pure, and work quickly with the sponge and lightly, and you will make the paintings appear as newly painted; if you wish to clean them, take white of egg and fig’s milk, mix well as you know how, and then proceed.”

Ignatius Hugford, Madonna and Child; imitation of a thirteenth century panel, 1758. Impruneta, Pieve di Santa Maria.

Ignatius Hugford, Madonna and Child; imitation of a thirteenth century panel, 1758. Impruneta, Pieve di Santa Maria.

To do the same:iv “Take one pound of the finest white soap, two ounces of ammonium salts, three ounces of white cherry gum, as much well water as is required, let everything dissolve thoroughly and then use as above.”
Or else: To clean figures painted on a wall or on a panel so that they appear new : “Take oak ash and the same quantity of quick lime and mix everything together, making a warm lye. Then take some honey, black soap, yolk of egg (equal quantities of each), and make sure that everything is well bound together, then rub with this tempered mixture which is proven [to work]; if the figures are painted on wood, only the mixture is necessary, and then to wash them with a sponge”. Naturally, one or other of the recipes would have been chosen according to what was compatible with the technique of the figures requiring cleaning (painted or sculpted), and also according to the speed required for the results. A slower and less hazardous method might be chosen not only to avoid the risk of damaging an important painting but also, from a more craftsman-like point of view, in order to save on the time which the repainting of an abraded area would have taken, with all the added problems of making new parts fit in with the old.
It is from precepts such as these that we can better understand what happened at the 14 beginning of the sixteenth century to Piero della Francesca’s San Jerome, now in Berlin. Up to the restoration carried out between 1968 and 1972, the presence of sixteenth-century paint covering much of the little panel which was dated 1450, had been noted, but it was thought that the work that had been left unfinished had then been completed at the turn of the sixteenth century. The highly abraded paint surface that emerged on removing the overpaint made it clear that the repainting had been a means of correcting the excesses of 15 the cleaning; repainting showing great respect towards the figure, as well as the niche carved out of the rock containing books and the cartouche with the signature, but more freedom in the landscape, and with the trees “restored” exclusively to fulfil their icono- graphic function. The most evident stylistic updating (which led to misunderstandings by Bode, and a doubt on the authenticity of the work on the part of Longhi) can be seen in the sky, in which the clouds that articulated the depth of the painting (as they do in the Baptism in the National Gallery, London, and in The Battle of Constantine, Arezzo) were painted over. The author of the repainting replaced the highly developed spatial definition of the sky painted by Piero della Francesca, which was typical of fifteenth-century taste, with a sfumato which, following the norms of early classical taste, from the intensity of the far sky becomes paler by degrees towards the horizon.

Circle of Guido da Siena, Saint Dominic; palimpsest panel, Cambridge (Mass.), Fogg Art Museum.

Circle of Guido da Siena, Saint Dominic; palimpsest panel, Cambridge (Mass.), Fogg Art Museum.

More often, it is only a question of a change in taste that leads to modifications in a painting and to the elimination of some detail that is disturbing because of its antiquated appearance: Michiel, for instance, when describing a Giovanni Bellini in the house of Antonio Pasqualino on 15 January 1532 (it is difficult not to think that he is referring to the 18 Madonna Frizzoni in the Museo Correr in Venice). He notes the repainting by Catena, observing the pictorial merits of the work, but without passing comment on the problem of different hands:
“The half-length figure of Our Lady, much smaller than life, painted in glue- size (a guazzo), was by the hand of Giovanni Bellini, reworked (riconciata) by Vincenzo Catena, who replaced the textile in the background with a blue sky. It is many years since he did this and it is clearly outlined, with the strong highlights
poorly blended with the half-tints; nevertheless it is a work worthy of praise for the grace of the heavens, for the drapery as well as the other parts”.21 Repainting could aim to give a more traditional air to a painting, as well as updating it; this can be seen very well in the Orazione nell’Orto in the predella of Raphael’s Colonna Altarpiece, which was entirely in keeping with the reputation of the “simple and venerable” women, as Vasari described the nuns of Sant’Antonio in Perugia, for whom it was painted. A comparison with the small cartoon of the Pierpoint Morgan Gallery and with the X-radiograph shows that the Orazione dell’Orto is rendered easier to read with the flying cherub replacing the original chalice, and that greater impact is given to Christ’s profile, which originally was too foreshortened for the image to be read easily by an uneducated eye. The small adjustments are the work of an Umbrian painter, contemporary of Raphael, rather than by the master himself. However, it is not always easy to distinguish between autograph alterations and those carried out by an assistant, a follower or a trusted collaborator of the artist. I am thinking of works that remain in the artist’s studio for a long time and are picked up again at a later date, as was the case with Correggio’s unfortunate Madonna del Coniglio: all the final adjustments, 19, 20 made by the artist himself, were removed during the wretched restoration of 1935.
There are times when the alterations or reworkings revealed by the X-radiographs, ultraviolet or infrared examinations, cannot be interpreted except as being by the hand of the artist, the author of the original version. In order not to destroy these autograph re-elaborations, it would suffice, as a philosophy of conservation, to remember the principle that does not allow for restoration (overpaint) to be removed unless it can be proven to be such through dating. I am thinking, again, of a particular example: Paolo Uccello’s Madonna and Child in the National 16, 17 Gallery of Ireland, Dublin, in which the removal of the blue veil covering the Virgin’s head attenuated the look of a “Peasant Virgin”, to use Alessandro Parronchi’s vivid expression. This, on its own, can explain why either the commissioner, or Paolo Uccello himself, had opted for a solution in which the Virgin’s head played a less active role within the character of the image, which otherwise has more in common with a perspectival theorem than an object of devotion.

Head of Livia, reused in the Herimankreuz; Cologne, middle of the eleventh century. Cologne, Diocesan Archiepiscopal Museum.

Head of Livia, reused in the Herimankreuz; Cologne, middle of the eleventh century. Cologne, Diocesan Archiepiscopal Museum.

Although the repainting covered some areas of damage, a later date is not possible because no-one would have taken the trouble to imitate the style of Paolo Uccello (even allowing for recognition of the authorship of the panel, as well as the ability and the understanding to do so), at a time when repainting alterations were still largely a matter of bringing works up to date. The veil that was removed shared precisely those stylistic characteristics that many prefer to attribute to the “Maestro di Prato”, as distinct from Paolo Uccello. At the end of the nineteenth century, when the little panel first entered the art market, we find it in the Bardini Collection with an attribution to Lorentino di Arezzo; such a solution, already confirmed by the earliest photographic images, would never have been considered at a time pre-dating the insertion of the panel within the works of either Paolo Uccello or the “Prato Master”.
At times it can be a much more straightforward case of taste dictating an autograph repainting, as in the case of the portrait by Antonello da Messina in Berlin, once upon a time 21 dated 1478. We are indebted to Longhi for identifying the sky and the landscape as old repainting: to the stylistic elements (“the projected shadows in the head are too strong to have been conceived in ‘plein air’ ”) are added the optical behaviour and physical characteristics of the paint with which the sky and landscape are rendered, typical of overpaint. They lack luminosity, because of the original black paint that continues to show through, whilst “to make a bit more room for the landscape (nevertheless somewhat curtailed) the right shoulder has been dismantled so that it no longer has the perspectival weight found in the master’s other portraits.” However, on close examination of the painting, one can detect that the lead-white or lead-tin yellow highlights on the hair are in fact painted over the sky; despite its incongruities, the repainting is clearly autograph. Should we want to express our dissatisfaction, we will have to limit ourselves to suggesting Jacobello da Messina as the author of the landscape, or else find other solutions within Antonello’s workshop.
In the light of what in the idealist theory of restoration is called the “aesthetic case”vi (“istanza estetica”), the conservation of a reworking executed by the artist himself (or within his workshop) is not a foregone conclusion, as the original painting often has an effect of greater freshness, whilst the reworking will have the material dullness, the opacity, characteristic of overpaint. It is, I believe, giving way to temptations of this kind, that in 1967 the lunette forming the pinnacle of the Pala Felicini by Francia (Dead Christ with Angels) was returned to its initial state, which is dated 1494. This involved the removal of the complete repainting of the work that was executed in the later style of the painter, probably towards 1515, when other adjustments were also made to the central section of the altarpiece as can be seen in the cleaning test on the right shoulder of Saint John the Baptist. If the quality of Francia’s earlier painting aesthetically justifies such a debatable course of action, one must nevertheless note that the solvents and reagents used to remove the second painting did cause an impoverishment of the 1494 paint layer, with some areas of damage that are now particularly noticeable in the most visible wing, belonging to the angel on the right.

Giotto, Baroncelli Coronation; polyptych reframed by Domenico Ghirlandaio, Florence, Santa Croce.

Giotto, Baroncelli Coronation; polyptych reframed by Domenico Ghirlandaio, Florence, Santa Croce.

The restoration of frescos
The problem of the conservation (manutenzione) and the substitution of cycles of frescos when these have reached an extreme degree of disrepair can be followed clearly in the Venetian documents of the Palazzo Ducale from as early as 1409, when it was noted that in the Sala del Maggior Consiglio “much has been lost in the paintings”; however, the Ufficiali al Sal “have [had] these paintings restored with the money of the community, which could be done for an expense of less than two hundred ducats”.vii A deliberation of the Senate of 9 July 1422 reiterated the necessity of entrusting a painter with the upkeep of the paintings which “fall apart as we speak”, “for the everlasting glory and praiseworthiness of such solemn, ceremonial works and for the honour and glory of our lord and city”.viii On 1 September 1474, Gentile Bellini’s offer was accepted, to maintain the paintings of the Sala del Maggior Consiglio “in good condition” (“ben in conzo”) without other remuneration than the concession of the first available brokerage of a warehouse (“sanseria de fondego”). And, in 1479, when Gentile was in Constantinople, the task was entrusted to Giovanni Bellini.

Paolo Uccello, Madonna and Child; after the removal of the autograph repainting. Dublin, National Gallery of Ireland.

Paolo Uccello, Madonna and Child; after the removal of the autograph repainting. Dublin, National Gallery of Ireland.

Paolo Uccello, Madonna and Child; with the autograph repainting. Dublin, National Gallery of Ireland.

Paolo Uccello, Madonna and Child; with the autograph repainting. Dublin, National Gallery of Ireland.

When, in 1515, Luca Luchini (assistant to the deceased Gentile and to Giovanni Bellini) offered his services for the maintenance of the paintings in the Sala del Maggior Consiglio, he was no longer referring to the late Gothic frescos, but to canvases painted by the Bellinis, by their assistants, and to the canvases by Alvise Vivarini and Carpaccio: there came a moment in time when, to preserve the order in the hall, it became necessary to replace the paintings which were too damaged. In this instance, the move was also made from fresco to canvas. In Lorenzo Malipiero’s Annali Veneti, Gentile and Giovanni Bellini were not referred to as having painted anew, but as having restored the Battaglia dei Veneziani con Barbarossa; this concept of restoration would mean that still today, in the canvases by Tintoretto, Bassano, Veronese and his heirs, we would be seeing the same hall as originally decorated with the frescos of Gentile da Fabriano and Pisanello, at least inasmuch as they celebrated the same episodes of Venetian history. Or again, in the Sala dei Giganti di Liviano in Padua, we would have before our eyes the cycle of Illustrious Men (Uomini Illustri) which the Carrarese commissioned in the fourteenth century, in the frescos renovated by Domenico Campagnola and his assistants in 1540.
The making good of lost parts, or the renewing of areas of perilously attached intonaco, necessarily involved interventions that one could class as restorations because of the need to harmonize the new areas with the original, using techniques adapted to match the original paint application, and imitating some of the stylistic features, even if not showing a real understanding of the style. In 1467, when Benozzo Gozzoli worked on the Maesta by Lippo Memmi in San Gemignano (and this is after Bartolo di Fredi’s additions in the second half of the fourteenth century), he imitated the pointed feet of Trecento painting in the figure of Saint Louis, and reproduced the friezes and the uncial script (already obsolete) of the original inscriptions with great care.
Analogous, but also more complex in his capacity to achieve a balance between his own style and that of the original, is the manner in which Pietro di Giovanni Orioli carried 22 out the vast integration on the extreme right of Ambrogio Lorenzetti’s Buongoverno. Until the recent clarifications on the identity of Orioli by Alessandro Angelini, the portion repainted in 1492 had been confused with the later restorations by Girolamo di Benvenuto: the earlier restoration follows the stylistic characteristics of the fourteenth-century master in the depiction of the mountains, especially in the way that they are profiled along the horizon against the ultramarine of the sky. More than one detail bears comparison with the landscapes by Orioli, for instance the Sulpicia in the Walters Art Gallery, Baltimore, which, despite some hesitation in the handling, eliminate any possible doubt on the authorship of the integration in the Buongoverno .

A History of the Restoration

Antonello da Messina, Portrait of a Young Man. Berlin Dahlem, Staatliche Museen.

Beato Angelico, Saint Mark; after the removal of Lorenzo di Credi’s repainting. Chantilly, Musée Condé.

Beato Angelico, Saint Mark; after the removal of Lorenzo di Credi’s repainting. Chantilly, Musée Condé.

During the Renaissance, accounts have also reached us of the transfer of frescos: Michiel in the house of Alessandro Cappella in Padua recalled seeing “the head of Saint John in fresco on wall, by the hand of the Florentine Master Cimabue, now placed in a wooden frame, which was removed from the church of the Carmelites when it was burnt down”. This is one of many examples of the removal from a fresco of small sections of the intonaco, such as we encounter in all periods, and which were carried out when demolition or whitewashing was impending. Often, the loss of a head in a fresco revealed when the whitewash was removed indicated that an attempt had been made to remove such painted “crusts”. There are countless examples, right up to Stendhal writing in 1837 about the Papal Palace at Avignon, where the occupying soldiers had been busy cutting out “heads by Giotto” from the fourteenth-century frescos, in order to sell them to passing tourists.
Quite other is the aim of the trasporti a massellolx of wall paintings. The earliest such transfer for which we have documentation is that of Piero della Francesca’s Resurrection in Sansepolcro, transferred from one wall to the other of the Palazzo Comunale, presumably around 1474, when building work was being carried out. The surface is well preserved, the only verifiable damage being in the loss of part of the columns which framed it in perspective.
That both Vitruvius and Pliny the Elder should refer to painted walls being moved from Sparta to Rome in 59 BC probably made of the operation yet another example of aparagone in which the architects of the Renaissance pitted themselves against the Ancients. In a manuscript that can be attributed to Baldassare Peruzzi, in which he deals with the merits of brick walls (which is exactly the same context in which Vitruvius speaks of the transfer of mural paintings), Trasporto or stacco a massello is a different method of detaching frescos from their original walls, as the painted plaster is removed along with part of the wall. With the technique of strappo, only the painted “crust” is removed.
He noted: “In my city Siena, a work worthy of great admiration, eighteen foot across and twelve foot high, of wonderful and ancient paintings, cut out and transferred to a new location in the residence of the Merchantia.”32 From the fifteenth century onwards, the transfer of frescos a massello was practised, and destined to have a long life right up to the attentions devoted to it by Forni in 1866, even if the practice did not always have the felicitous results we can see in Piero della Francesca’s Resurrection; it was Mengs who made the point that: “in similar cases, it always happens that with the fresh damp, and with the salts in the plaster, that a kind of encrustation forms, covering the frescos and making them apparently disappear”.