Lute of the Month



The Lute Player, 1596, Wildenstein Collection

Over the past twenty years historians of early musical instruments generally, and lute makers in particular, have become heavily involved in the study of musical iconography as a necessary tool to make sense of the complex development of different instruments as they changed through the centuries in different regions of Europe . A certain expertise has now been built up as, increasingly, correlations are established between contemporary descriptions of instruments, the requirements of the available music, the history of varnish and string technologies, the treatises on playing techniques, the surviving instruments and their contemporary depictions. More and more the iconography has become the matrix into which other components must fit. It has forced us to realise how few surviving lutes are what they at first appear to be; how many have been altered, often several times, to meet the changing musical needs of the day. How many, indeed, in museums throughout the world, are simply impossible instruments confected by the large 19th century faking industry. In all this research art history has been invaluable and every good lute maker now spends almost as much time in galleries looking at paintings as in museums looking at the few surviving instruments. But of course occasionally this specialist interest enables us to repay the debt and see aspects of paintings which are not apparent to the untrained eye, even of an art historian.

So, as a lute maker, I was very interested to hear that a new Caravaggio portrait of a lute player had been discovered. Caravaggio was so exact a realist, so careful about the details of instruments and worked in Italy at such an important period for the history of the lute that any new picture of his promised to be both fascinating and potentially revelatory. When I first looked at a poor quality B & W reproduction in a newspaper, however, it was immediately apparent that there were problems with this picture, and now that I have had the opportunity to examine the very good reproductions in the exhibition catalogue [‘A Caravaggio Rediscovered’ Keith Christiansen, Metropolitan Museum of Art, New York,1990.] it seems to me that there are so many anomalous details that they ought to be aired and, hopefully, cleared up one way or another.

In general what the new picture shows is clear enough. It is a version of the Hermitage Lute Player in which the six course lute there shown has been replaced by what appears to be the seven course lute out of the “Musicians” painting in the Metropolitan Museum, complete with its black pegs, ebony fingerboard, ebony edging to the soundboard and ebony strips between the ribs of the back. The shape of the pegbox too with its wider end is similar to the lute in the Metropolitan painting.

However I was immediately struck by the oddity of the visible end of the lute bridge.

The Lute Player, Detail, Wildenstein Collection

This is quite unlike normal Italian lute bridges of this period, if anything it looks like a back-to-front mandora bridge from the 18th. Century.

Mandora based on Sympert Niggel, 1754,

In the Hermitage picture less of the bridge is visible but the end does clearly turn towards the top of the instrument in the way one would normally expect, compare for instance the lute by Georg Gerle

Georg Gerle, Kunsthistorisches Museum, Vienna, No.A 35

and the bridge in the Berlin picture, which is very clearly shown

Love as Conqueror detail, c.1602, Staatliches Museum, Berlin

Alerted by this anomaly, one looks more closely at the bridge, for after all maybe this is just the sort of revelation that changes our ideas about 16th. Century lute construction. But no, looked at closely, the whole treatment of this new bridge betrays complete uncertainty about what the actual nature of the structure is. The exhibition, perhaps inadvertently, may have led people to overlook this particular detail because the Sixtus Rauchwolff lute, which was displayed as an example of a typical renaissance lute, has a replacement 19th Century guitar bridge, of which the outer part does indeed curl backwards, somewhat in the manner of the painting

Sixtus Rauwolf, 1596 Metropolitan Museum

Although the refretting and restringing are remarked on in the catalogue this anachronistic bridge is not mentioned.

Next, consider the black edging to the soundboard, this appears to be edgeways on, round the outer edge of the soundboard rather than let in flat to the top surface of it.

The Lute Player detail, Wildenstein Collection

This construction, if true, would be unique among surviving 16th and 17th century lutes or even, to my knowledge, among lutes shown in paintings of the period. Superficially the “Musicians” lute has a similar edging but the position of the line of reflected light

The Musicians detail, Metropolitan Museum, New York

shows that the wider surface is in the same plane as the top of the soundboard, which is the normal form of such an edging, not at right angles to the soundboard, as appears to be the case with the new lute. The painter of this new lute seems to have thought of the ebony edge as a kind of capping strip round the outer edge of the belly; if indeed he has thought at all, for again the overwhelming impression is of lack of close observation of the structure. Compare this quality of observation with the treatment of the edge of the soundboard in the Hermitage lute,

The Lute Player detail, c. 1600, Hermitage Museum

here indeed there is no edging at all (which was perfectly normal for 16th. Century lutes) but look at the beautifully observed worn and shiny edge to the wood of the soundboard, even the grain structure of the wood is clear. Any player of the lute would recognise that appearance, and it confirms Caravaggio as a master of close attention to the texture and structure of things. Even in a picture such as the Musicians the looser brush strokes and freer technique do not conceal the essential clarity of his gaze; where details are suggested they clearly represent actual structures rather than generalised approximations.

There are 14 pegs which, for a lute of this period, has to imply a seven course instrument with a double top course, indeed this would be the most usual string layout for an Italian lute at this time. However there is only one top string present and, even more suspiciously, only one groove in the nut, even though there is a broken string dangling from the second peg. ( Incidentally the catalogue is quite wrong to suggest that such dangling string pieces are, as it were, spares in case of breakage [Catalogue p.86]; broken strings cannot be knotted and then reused, they would be completely untrue and out of tune. In most paintings broken strings have symbolic connotations, either of mortality, as in Holbein’s Ambassadors [see essay by Mary Hervey pub. 1900] or of social or moral disharmony.) While the Metropolitan picture has too much damage in this area to be sure about the stringing, the Hermitage lute clearly has a double top course to go with its twelve pegs. Furthermore, when one examines the strings in the pegbox of the new lute it becomes obvious that the pegbox has been strung up completely out of order. This is an important matter for a player who must know, without looking, which peg controls which string. Important too, one supposes, for an artist who is looking at an actual instrument and working in the service of a patron who is a lover and collector of musical instruments. But here we have a fourth course whose two adjacent strings go to the seventh and twelfth pegs, and fifth, sixth and seventh courses which vanish into the pegbox at angles which make it impossible that they are heading for any particular peg.

The Lute Player detail, Wildenstein Collection

This might seem to be an insignificant matter but is one which the painter of the Hermitage lute has bothered to get right. Once again it seems as if the painter of this new lute has not really had his eye on the object. Can this really be by an artist famous for painting every detail from the life?

The rose, or soundhole carving, is also not even hinted at in this new picture, just a generalised dark patch on the surface of the soundboard. As this is such a prominent feature of a lute it is almost universally attempted in pictures of the period, for instance every other painting in the catalogue showing a lute being played gives a more or less exact account of the lute’s rose. Only the new picture chooses to generalise in this way. Once again the Hermitage painting has a beautifully observed passage of carving in the rose

The Lute Player, rose detail, c.1600, Hermitage Museum

Turning now to the position of the player’s hands; it was the usual practice to place the little finger of the right hand on to the soundboard as a steadying and locating measure while playing. Indeed there are often wear marks on the soundboards of surviving lutes which attest to this practice and alert us to changes in playing positions through the centuries. The Hermitage picture shows just the sort of position one expects to see, with the little finger stretched out and the other fingers in various stages of preparing for, and recovering from, plucking the strings; the little finger is not quite touching the soundboard which gives a poised sense of movement, of actual playing going on at the moment which has lifted the player’s hand for an instant. The new picture, however, is a much more earthbound affair; the player has his third and fourth fingers firmly on the soundboard in contravention of all the contemporary instructions, and is clearly plucking the strings from beneath in just the manner of a modern guitarist. Moreover a badly taught modern guitarist with his top knuckles collapsed inwards! In fact when you examine this hand closely it is a very peculiar affair indeed, the perspective and scale of the fingers does not match up with that of the body of the hand and the wrist, producing a kind of deformed claw, quite unlike the elegance of the Hermitage hand. Maybe this is part of the “north Italian empirical approach to spatial projection”, to which I shall return later.

Looking now at the violin:

The Lute Player violin, Wildenstein Collection

this appears to be a compilation, taking features from each of the four violins shown in other Caravaggio paintings. The little inlaid or painted decorations on the corners come from the Rest on the Flight into Egypt

The Rest on the Flight from Egypt detail, 1597, Galleria Doria-Pamphili, Rome

The unusual scroll with its single fluting is also featured in Love as Conqueror

Love as Conqueror detail, c.1602, Staatliches Museum, Berlin

and The Musicians

The Musicians, Metropolitan Museum

The Hermitage Lute Player violin has the, to our eyes, more usual post-Amati double fluting. The curious curve to the top of the pegbox cheeks, which seems to threaten almost to cut the head from off the neck, is also shown in Love as Conqueror, as is the ebony tailpiece and the strange bridge. The neck root, which joins the body of the violin without curving downwards to form a button at the top of the back, although unusual for post-Amati violins, does seem to be a feature of some very early violins and shows all these violins to be looking backwards rather than forwards for this date. There is a Linarol violin of 1581 in Vienna Kunst Historisches Museum [No. C 96, illustrated in Anthony Baines, European & American Musical Instruments, London,1983 Pl. 13 & 14] which has just such a single flute scroll and neck root to a slight point on the back.

However what really does look odd to an instrument maker is the grain of the front of the violin

The Lute Player violin, Wildenstein Collection

it is wide and coarse to such an extent that it appears out of scale with the instrument as a whole. Wood with grain this wide is rarely used on violins, even very cheap ones. The other major problem with the violin is the arching of the front in the top bout, the top right hand side nearest to the spectator. It slopes down towards the edge in a convex curve which is quite uncharacteristic of normal violins, which usually have a concave section along the line of the purfling before rising toward the rim of the final edge. While one should not dogmatically say that such an arching is impossible for a violin of this period it is so very defining a characteristic of the violin as an acoustic structure that it does cast doubt once again over whether the painter of this picture had an actual violin in front of him at the time. As with the lute, we need only to turn to the Hermitage painting to see a beautiful and fully convincing picture of an early violin with the typical arching one would expect.

The Lute Player, violin detail, c.1600, Hermitage Museum

The perspective of the Wildenstein violin in relation to the table is also particularly awkward, the instrument appears lodged into the surface of the table in an almost surreal manner. It makes sense when read with the figure but not when read with the music and the table top.

Next there is the question of the violin bow. Bows of this period had clip-in frogs which were pushed into a little recess in the bow stick in order to tension the hair which was fastened firmly in mortices at the tip of the bow and just behind the frog. This system requires that the frog be a separate piece of wood from that of the stick itself, as shown in the Hermitage bow, where the line between the two pieces of wood is clearly visible.

The Lute Player, bow frog detail, c.1600, Hermitage Museum

In the Wildenstein picture, however, it is not at all clear where one piece of wood ends and the other begins, one seems to flow into the other suggesting it is a fixed frog.

The Lute Player bow frog detail, Wildenstein Collection

This was the earlier system and is shown with great clarity in the painting by Costa in the National Gallery [Catalogue p 22, Fig.10]. However this system requires that the hair be tensioned at the tip like an archery bow which is exactly what the Costa painting shows, in its untensioned position. The “new” Caravaggio bow however has an impossible amalgam of both systems with a fixed frog and hair fixed at both ends.

Having been alerted by these problems with the instruments, which is the area of my own expertise, I was prompted to examine the rest of the painting with a more jaundiced eye and here too there seem to be anomalies. The first difficulty, even apparent from the B & W reproduction (and perhaps more apparent there without the distractions of colour), is the perspective of the table. It is clearly seen from a higher vantage point than the figure itself and when the two are read together the table appears to slope sharply downwards towards the viewer. So much so indeed that one fears for the safety of the recorder in the foreground.

The Lute Player, 1596, Wildenstein Collection

Compare this with the Hermitage

The Lute Player, c. 1600, Hermitage Museum

and the Metropolitan pictures, where the horizontal surfaces are obviously from the same vantage point as the other figures. Indeed the use of such a horizontal surface carrying instruments is almost a cliché for pictures of this period and one would have expected that such a master of perspective as Caravaggio would have been able to handle an elementary matter like this a little more convincingly.The catalogue refers to this rather coyly : “ However it is worth noting that there is no single vanishing point: the side wall is projected from a different position. In this respect Caravaggio remained faithful to a north Italian empirical approach to spatial projection.” [Catalogue p.60] Since earlier in this section it had been argued that this version of the picture is later than the Hermitage version, in which a unified perspective is used, it would be less confusing, but perhaps more revealing, to say that with this picture Caravaggio returned to an “empirical approach”. But where would that leave the earlier suggestion that “Caravaggio’s interest in perspective was evidently stimulated by his residence with Del Monte, whose brother was then writing a treatise on the subject,....” ?

Does this mean that Caravaggio first painted the Hermitage picture using a modern-seeming unified perspective, afterwards to have his interest aroused in perspective theory by Del Monte’s brother in order to return to a more primitive type of spatial projection (which he had never previously used) just for this one picture; and afterwards to paint with a unified perspective for the rest of his life? It is an interesting thesis.

In his other paintings he seems almost to have enjoyed setting himself problems of foreshortening and perspective in order to solve them with panache. See, for instance, the jumble of instruments in “Love as Conqueror”

Love as Conqueror detail, c.1602, Staatliches Museum, Berlin

Secondly, where is the light source which illuminates the right-hand side of the player’s left sleeve? One of the characteristics of Caravaggio’s theatrical style is the single light source which throws an often low raking shaft of light into the circumambient gloom, leaving deep shadows and showing the forms in high relief. Here, however, while the main structure of the picture follows this schema, shown in such clarity by the Hermitage painting, the details of the cloth of both sleeves reveal an anomalous second light source to the right of the subject. This is especially obvious on the extreme right of the sitter’s left sleeve, where, instead of graduating round into the surrounding darkness, there is a sudden change in illumination and the cloth ends in a surprisingly harsh bright contour against the dark background. Unaccountably, other objects in the painting do not pick up this light source. Even more surprising is the private local light source which illuminates the extreme right hand side of the sitter’s right hand sleeve where it rests against the front of the lute. This sleeve manages to cast a deep shadow on the lute soundboard but is itself glowing with a strong light, seemingly from the lute. Meanwhile the hand and wrist of the player, although shadowing the soundboard, cast no shadow at all on the sleeve! Is there perhaps some Transylvanian influence here?

The Lute Player, 1596, Wildenstein Collection

Which leads on to the more subjective matter of the deeply unpleasant looking face. Maybe in the intervening time the sitter had taken to wearing makeup but the contrast with the Hermitage face could not be more extreme, given that it has more or less the same features. The effect reminds me of the description of Aschenbach in Thomas Mann’s Death in Venice. It is true that some of Caravaggio’s faces do seem to be a kind of mask worn by characters in a play; in Shakespeare’s telling phrase,”They are the lords and owners of their faces” which seems to speak of a distance between the real self and its persona; but nowhere else in his work have I seen such a cardboard cutout as this. The flesh tones are flat and lifeless. The eye rests on the surface and does not look through to feelings about the sitter; there is, as it were, no spiritual light source illuminating this part of the painting.

In view of all these problems I feel that m’learned friends would say that the attribution to Caravaggio was unsafe and unsatisfactory.

And that, m’Lud, concludes the case for the prosecution.

For those interested in the case for the defence, the Metropolitan Museum have now offered a free download of the catalogue I refer to in this article.

© David Van Edwards, Norwich 1996 and then published in Lute News No. 50 June 1999.

Since this was written there have been further developments and another much better version has come to light, which will be featured on the cover of the December 2015 copy of Lute News

Index of other iconography essays

History of the LuteBuild your own
Baroque Lute
Build your own
Renaissance Lute