RECONSTRUCTING AN INSTRUMENT AFTER A PORTRAIT BY JAN VAN HEMESSEN, ANTWERP, ABOUT 1530
The van Hemessen Clavichord
Most of what might be considered the major schools of keyboard-instrument making have surviving examples. Although the instruments from earlier traditions are often altered – and here one immediately thinks of the sixteenth-century Venetian harpsichords, or the majority of instruments from the Ruckers family – there is still an existing group of instruments from which conclusions about their original state can be reached.
The early-Flemish clavichord school is not represented by any surviving instruments, but by a group of illustrations – both paintings and drawings. These images show that there was a variety amongst the details, but all of the examples shared the same main characteristics. These features are also found in the slighter-later surviving virginals by Ioes Karest – one of which shall be discussed in some detail shortly – and in other paintings of virginals from the period. Although no Flemish clavichord from the period survives, there are two other instruments, which I shall briefly discuss later, that have some of the same characteristics, even though they were made perhaps a century after the period discussed here.
The origins of the early-Flemish clavichord school can be found to have been established by around 1475, at which time a painting by the “Master of the Duke of Burgundy” shows a clavichord played by an angel. The remainder of the illustrations cover a period from circa 1500, to around 1600, although the majority are from the period from about 1525 – 1550. When the instruments appear in paintings they are found in only two contexts. In the first they are seen in scenes involving angels, and in the second they are used as part of a highly symbolic group in which a young girl is playing or carrying a clavichord in a room in which an ointment jar – a symbol of Mary Magdelene – is also prominently displayed. It should also be briefly mentioned here that very similar instruments are occasionally found in paintings from outside the Low Countries, for example in the painting attributed to Gaspar Vaz or Nasco Fernandesz which can be seen on the alter in the Monastery of São João de Taroucha in Portugal.
The finest of these paintings is almost certainly that by Jan Sanders van Hemessen. From the research carried out by Burr Wallen as part of his PhD on van Hemessen, submitted to New York University in 1983; and in the article by Wallen in the Macmillan-published The Dictionary of Art, it is known that van Hemessen was apprenticed to Hendrik van Cleve I in 1519-20, and possibly worked with the “Master of the Magdelene Legend” at Mechelen. (By way of a brief clarification, I should say that Mechelen is a town a little north of Brussels which was a royal court, both then and now well-known for its lace-making and as a major ecclesiastical and that otherwise Anonymous artists to whom a group of paintings can be attributed are often known as the “Master of”, so that, to use an English example so not to cause any confusion with the subject matter here, the “Master of the Countess of Warwick” is an anonymous mid-sixteenth-century artist who was connected to, and is in modern research associated with, the Countess of Warwick of that time). From about 1525 van Hemessen appears to have been based at the court of Margaret of Austria and Mary of Burgundy at Mechelen. Here he would have come into contact which other artists such as Jan Gossart and Jan Corneliesz Vermeyen, who was in charge of the court’s workshop at that time. An artist (or group of artists) known as the “Master of the Female Half-Lengths” was also probably at the Mechelen court in the 1520s and 1530s.
The van Hemessen clavichord painting is the only one for which – intentionally or otherwise – the sitter can be reliably identified. She is believed to be Margaret of Parma, who was the illegitimate daughter of Emporer Charles V. Margaret was born in 1522, and the attribution of her as the artist’s sitter comes from the striking facial resemblance to a slightly earlier painting of her by Jan Gossart, and a record of three paintings to be made of her in 1528-29. Two of the paintings survive – this clavichord portrait, and another in which she is weighing gold. The third, known only by a later copy, shows her writing. No doubt the paintings were commissioned to show the education and accomplishments of the young girl, reasonably certainly with a view to being shown to any royal family of Europe who might have a son in need of a princess. There are indeed records of payments to show that Margaret had music and dance lessons at this young age.
* * * * *
Using a two-dimensional painting in order to reconstruct an otherwise non-existent musical instrument is at best a speculative academic exercise and at worst entirely pointless. No matter how precise the illustration might be, there are always some parts which are hidden from view which present interpretation problems for which other, equally-valid, solutions might be found. However, I personally believe that if a reconstruction of a musical instrument has been achieved on paper, it must then be actually made in order to learn something tangible from the process. That said, the first part of such an exercise must start with turning a two-dimensional representation of a three-dimensional scene into views from other angles – principally in this instance a top, front and side view, from which an instrument can be constructed. I should point-out, and emphasise, that my purpose with both making a drawing and an actual instrument has been to reconstruct what the artist painted as far as possible, and there has never been any attempt to alter dimensions in order to follow any putative design-method that has been proposed for other instruments by writers such as Birkett and Jurgenson, Henkel, Heyde, Koster or O’Brien. Certainly there are some results which imply interesting proportions – for example the bridge to right-case-end is one-ninth of the total case length, and is directly behind the right keywell side – and if others wish to pursue this approach they are welcome to my raw data.
In making the drawing, the initial starting point was to set some basic parameters from which the details could then be extrapolated. In this instance the keys formed an ideal starting point for which a nominal “unit measurement” – essentially an instrument inch – could be measured, and then the overall case length, width and depth could be established. This was initially done by making a simple cardboard mock-up and trying to recreate the artist’s position, and then confirmed using a 3D CAD programme. Once these basic dimensions were established in three directions it was then possible to proceed with other measurements.
The low angle from which the artist has depicted the instrument, allowing the sitter to look down at the face of the viewer, means that a number of parts of the instrument are greatly obscured. The method by which the painting was turned into a plan-view drawing and the problems that entailed have been published elsewhere, and need not be repeated here, but there are several features which are worthy of mention here.
The majority of the keyboard can clearly be seen, and there is no doubt that the top of the compass goes to a’’, without the top g#. This treble extreme is found in a number of instruments from Antwerp and Italy, often in early surviving examples having a corresponding bass which descends to F with Bb as the first accidental. It can clearly be seen that Bb is, indeed, the first accidental, but the lowest note must be an E, rather than an F. This can be confirmed by counting the number of keylevers visible behind the nameboard, where we can see there are 39, one more than the number required for the common F,G,A – g’’,a’’ compass. The tuning of this lowest key was probably left to the discretion of the player according to the music being played. The angling of the keylevers behind the nameboard also suggests that the first fretted pair is between Bb and B, as would be expected.
There are a number of logical rules which must be followed without exception in order for the proposed reconstruction to be considered plausible, and to accept that the artist has accurately depicted the instrument. First, the bridge must be presumed to be perpendicular to the spine; second, the fretting pattern must be logical and based on historical precedent; third, the scaling must be reasonable and, although not necessarily Pythagorean, must not deviate too greatly; fourth, the bridge itself must be in a reasonable position; and finally, the projected tangent positions must be consistent with the splaying of the keylevers where they are visible in the painting.
These last three conditions are the most important in ensuring the painting can be considered as representative of a real instrument. The string lengths and projected bridge position must be worked out together from the evidence of the illustrated keylevers. In doing that part it became clear that the bridge position would be very close to the inside edge of the wrestplank. Using a three-octave span of 497 mm – chosen because it is found in the oldest surviving Flemish instrument, the 1548 virginal by Ioes Karest – there is only 29 mm from the edge of the wrestplank to the bridge.
At the time of making my initial drawing of the instrument I was forced to speculate that the only way the instrument could work would be if the wrestplank was not glued directly to the soundboard, using evidence from hitchpin rails of other early clavichords. Several years after presenting this supposition, an instrument in Stary Sacz, Poland, was brought to my attention through the initial work of Alicja Knast who kindly passed photographs of it to Maria Boxall who showed them to me. This instrument, although dating from 1634 – over 100 years after the van Hemessen – has exactly the gap between the wrestplank and soundboard that I had proposed. Although not Flemish, the Stary Sacz instrument is the only surviving example I am aware of which has much of the same construction features as the van Hemessen and other instruments from the early-Flemish clavichord tradition.
The making of the instrument is where, as mentioned previously, the true process of understanding the instrument begins. To give a single example, why should the artist have depicted the dovetails on the case corners unevenly spaced, so that the middle tail is closer to the lower than the upper one, and why are they reversed so the decorative part is on the sides and not the front? The instrument was made following what I believe is a plausible order and method of construction using, in this case, a single stick – approximately half-inch square in section and exactly the length of the instrument – which had all of the clavichord’s dimensions marked on it.
The instrument’s construction starts with the casework, and the above-mentioned dovetail question is one of the first that must be addressed. As part of the drawing process it became clear that the dovetails are either side of the wrestplank and hitchpin rail, which implies that the front and rear of the case have mortices into which the wrestplank and hitchpin rails slide, having already been glued to the case ends. This construction firmly locks the wrestplank and hitchpin rail in place so there is no chance of moving as a result of the tension of the strings. This not only explains why the dovetails are not evenly spaced, but because the wrestplank and hitchpin rail were already glued to – and were presumably also morticed into – the case ends – the ends must be attached to the front and rear, which forces the maker to use the decorative pins on the case ends. It also makes it clear that the instrument must be constructed from the top down, so that the rack (which I choose to varnish before cutting the slots) is glued next, followed by what we may refer to as soundboard liners – including those which separate the bottom of the wrestplank and hitchpin rail from the soundboard – before the soundboard itself is inserted. The soundboard barring is, of course, speculative, as is the question of a baseboard. The top of the soundboard is approximately halfway up the case height, and with the decorative cut-outs on the case ends and rear clearly visible it is obvious there is minimal space for a baseboard. Although there are comparatively few strings, there is nothing in the construction to stop the front and rear of the instrument being pushed apart by the string tension, and therefore it is almost certain there must have been a baseboard. Even with a baseboard of a similar thickness to the soundboard, it can be seen on the reproduction that there is considerable force trying to force the front and rear apart. To an extent this is unavoidable, given that the top of the soundboard is approximately one-and-a-half inches below the plane of the strings. It is possible to minimise any excessive risk of case movement by using some of the bars under the soundboard in a structural manner, having them glue to both the soundboard and baseboard, something done in my reproduction so that only the angled bar closest to the right end acts as a true soundbar, being glued only to the soundboard.
The soundboard itself can only be removed by taking off the baseboard and the liners below the soundboard, although apart from applying too much tension to the instrument from the strings it is unlikely to imagine a situation where removal would be necessary.
The bridge is also speculative to an extent. It obviously must be high, and the presumption is that it is not glued directly to the soundboard itself, but just sits on it, as occurs with early clavichords made in other traditions. In the reproduction I chose to have a cut-out under the centre part of the bridge, similar to what might be found on a viol, and the top of the bridge had an inlaid brass wire.
The other aspects of the instrument’s construction are not particularly associated with the early Flemish clavichord, although the style of lid battens which entirely close the end of the lid appears to be something of an early-Flemish characteristic. One of the great advantages of having a number of different paintings of instruments which are clearly built in the same tradition is that the different representations allow some of the speculative features to be based on other examples. For example – although the rack cannot be seen in the van Hemessen painting, it is visible in the Gossart example, and that image provided the basis of the design in the reproduction.
The instrument is strung in iron, with only the lowest four courses in brass. In fact, if the brass were taken any further up the compass it would break assuming that the treble was itself tuned near to breaking point. Needless to say, there is no way in which the pitch can be determined in any meaningful sense as there is no scalable item in the painting from which to draw any conclusions. Using the three-octave span of the Karest virginal the instrument plays comfortably at modern pitch or lower, and could go a semitone higher without any great risk of breaking strings.
One final aspect which should be commented on is the decoration. The use of maple for the case material appears to be a constant characteristic of the early-Flemish clavichord school, something which is not at all surprising given that it is also used in the earliest surviving Flemish virginals dated 1548 and circa 1550 made by Ioes Karest, and in Flemish paintings which include virginals from the period of around 1545 to 1565. It is also found in the earliest instruments built in Naples, including clavichords known as Numbers 2 and 3 in the Leipzig collection. This should not be too surprising as the Kingdom of Charles the Fifth went far beyond the Low Countries to include Spain and her Dominions, of which Naples was one. The painted-work as can be seen on the van Hemessen clavichord is found on various early-Flemish paintings and instruments, eventually forming the block-printed designs of the Flemish virginals and harpsichords of the late-sixteenth and seventeenth centuries.
Although no example of an early-Flemish clavichord has survived, it is very possible that an Anonymous instrument which is part of the Rodger Mirrey Collection, now part of the University of Edinburgh, represents the next stage of the Flemish clavichord, in which the soundboard is restricted to the right of the keyboard, and is positioned higher in the case, with a low bridge glued to the soundboard, but still having what might be considered the remains of the low soundboard under the keys, albeit one which in no way acts in that manner. It also has a slightly raised baseboard, a hitchpin rail which is not glued directly to the low soundboard, and a keywell which does not extend to the bottom of the instrument. Further, it can be seen from traces of wood still glued to the top of the lid that there was originally lid battens similar to those found on instruments of the early-Flemish clavichord school, and that there was paper covering the instrument. Although there are some clearly visible papers glued to the instrument at present, they are certainly later replacements, but small traces remain to show there were other papers very early in the instrument’s history, probably originally. All these features strongly suggest that the instrument was built in the Low Countries, probably around 1620.
Finally, I would like to make a suggestion about who may have made the clavichord painted by van Hemessen. There are obviously a number of potential candidates, but suggesting one over the others without any supporting evidence is of no value. However, there is a good case to be made for the Antwerp organ-builder Anton Moers. As discussed at the start of this paper, the portrait was painted in the Mechelen court by van Hemessen, as a group of three which were commissioned in 1528-29, of a girl – Margaret of Parma – who is known to have been given music and dance lessons. There is a Court record which shows that in 1529 – the same year as the painting – Anton Moers, a craftsman who had earlier built an organ for Mechelen, was paid for making a clavichord for the court. It does not seem too far-fetched to suggest that it was this instrument which was used in the van Hemessen painting.