Oxford Music Online
article url: http://www.oxfordmusiconline.com:80/subscriber/article/grove/music/12420pg2
Harpsichord, §2: The Renaissance
2. The Renaissance.
15th-century representations of harpsichords from various parts of western Europe generally
show short instruments with thick cases. Some do not appear to have a jackrail and may not have
worked by means of the standard jacks described in §1 above. Instead they may have had one of
the actions described and illustrated in the manuscript treatise of HENRI ARNAUT DE ZWOLLE.
Arnaut called his harpsichord the ‘clavisimbalum’. His design probably partly describes actual
constructional practice of the time; he may also have wished to give the clavisimbalum a
theoretical foundation based in geometry. This mixture of approaches resulted in some
inconsistencies. The design shows four types of action. The first and third are plucking
mechanisms that incorporate a swinging tongue that carries the plectrum, as in 16th-century
jacks; however, the part carrying the tongue is hung on an axle in a slot in the wrestplank (first
type) or is a large pivoting lever (third type). The small harpsichord played by an angel in
Manchester Cathedral, England (1465–8), is a convincing example of this type of jack action
without a covering jackrail. The second mechanism probably plucks, but without a swinging
tongue. The fourth (used in the DULCE MELOS) strikes an undamped string and was a forerunner
of the piano mechanism. Three of Arnaut's mechanisms are without dampers; this may have
been typical of 15th-century actions. Bird quill was probably used as a plectrum material, and
possibly also metal. Arnaut's design has a compass of B–a″ and can be made with either one or
two registers of strings; unusually, the second is aligned above the first, the strings being plucked
successively by the same jack. Arnaut's and other 15th-century harpsichords would have
sounded at a high pitch, from about a 4th to an octave above the 8′ pitch of the 16th century.
(i) Italy.
Although no 15th-century Italian harpsichords or representations of them are known to have
survived, it has been shown that Italian instrument makers were building harpsichords by 1452 at
the latest (Esch, H1979). Documents, manuscripts of Italian keyboard music (including organ
music) and intarsias suggest that a compass of FGA–g″a″ (or –c‴ or –f‴) was in use in the
second half of the 15th century. An intarsia of around 1520 in the choir-stalls of Genoa Cathedral
shows a single-register harpsichord with a compass of FGA–g″. Given the early date of the
intarsia and the compass, it is plausible that it also represents the type of harpsichord made in
the late 15th century. It has a bentside with two curves, a feature otherwise known only in virginals.
45 Italian harpsichords are known to survive from the period before 1590 – a greater number than
from any other region. Although there were some stylistic differences between harpsichords from
different towns on the Italian peninsula, broad similarities justify the term ‘Italian’, even though
there was no political unity on the peninsula until the 19th century. Slightly more than half of the
surviving 16th-century harpsichords were made in Venice. Guild regulations were less restrictive
there than in some places (e.g. Germany). To judge by the number of Venetian instruments that
made their way to other parts of Italy, the reputation of Venetian makers was considerable.
Alfonso II d'Este of Ferrara had at least six Venetian harpsichords, and Raimund Fugger (1528–
Grove Music Online
Harpsichord, §2: The Renaissance
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 1/9
69) in Augsburg had five. More 16th-century string keyboard instruments survive by Domenico da
Pesaro, who was active in Venice, than by any other maker, and seven of his 15 extant original
instruments are harpsichords. Also in Venice, Alessandro and Vito Trasuntino enjoyed good
reputations, and Baffo, Celestini and Francesco Padovano made instruments that show high
quality work. Most evidence of 16th-century harpsichord making comes from Venetian
instruments. Although several 16th-century virginals from Milan survive, no Milanese harpsichord
from this period is known. A group of harpsichords has been recognized as coming from Naples,
an important 16th-century musical centre. Harpsichords from Florence and Rome also survive.
The characteristically slender case shape of Italian instruments results partly from the practice of
doubling the string length at each octave down to f (sometimes to c), but also from the use of
longer bass strings than in other traditions. The case sides were usually not so deep as in
instruments from other countries. Little is known about the design and layout procedures used by
Italian makers. The string lines, plucking points and nuts (8′ and 4′) were sometimes marked on
the baseboard, but the fact that such marks are usually lacking suggests the use of standardized
designs and templates. Some 15th-century design traditions survived well into the 16th-century.
An early 16th-century Neapolitan harpsichord (no.175, Donaldson Collection, Royal College of
Music, London) has string lengths that double at each octave when measured between the
plucking points and bridge; this system corresponds to the design described by Arnaut de Zwolle
around 1440. The case proportions (excluding the visible part of the keyboard) of two octave
harpsichords by Domenico da Pesaro (1543, Musée de la Musique, Paris; 1546, Gesellschaft der
Musikfreunde, Vienna) are the same as that of the clavisimbalum that Arnaut described.
Harpsichord case-slides were usually thin (4–6 mm) and made of cypress, although maple was
occasionally used, particularly in Naples. Elegant mouldings at the top and bottom of the case,
typical of Italian harpsichords, have proved an important means of attributing unsigned work.
These thin-cased instruments were rarely painted but were provided with a separate, decorated
outer case, and are therefore often referred to as ‘inner-outer’ harpsichords. Supports for outer
cases survive in a variety of forms, some with simple, turned baluster legs, others carved, painted
and gilded. The thin cheeks at either side of the keyboard were reinforced by gluing on a second
piece of wood as thick as the case; these were then cut to scroll or other shapes, never being
being left square or slanted. Inlaid stripes of contrasting colours, forming geometrical patterns of
Arab origin, were used on the nameboard and the inside case above the soundboard in the best
Venetian instruments. Fine examples are the 1574 Baffo (Victoria and Albert Museum, London)
and an unsigned harpsichord (no.1883.718, Kunstgewerbemuseum, Schloss Köpenick, Berlin).
The nameboard, made of wood as thin as the case and removable, was sometimes panelled
with mouldings. The maker's name, if it appeared, was usually in small Roman capital letters.
Internal bracing usually comprised two to three stiffening rails nailed and glued to the bottom
boards; triangular blocks (called knees) maintained the sides perpendicular to the bottom
boards. One to three knees on the spine side and five to seven on the bentside was a common
arrangement. Since the case sides were thin they were glued to the sides of the baseboard for
rigidity, rather than to the top surface of the baseboard as in other countries. In a few south Italian
(probably Neapolitan) instruments the baseboard is about 5 mm above the bottom edge of the
case sides, presumably so that it is freer to vibrate. Diagonal struts from the bentside liner to the
bottom boards were also used, either with knees or, in some Venetian harpsichords, as the only
support for the sides. A decorative rose was often set into the soundboard, made usually of three
or four layers of thin wood veneer or sometimes of parchment, in gothic or geometrical designs. A
few instruments had three or four roses, echoing illustrations of 15th-century harpsichords from
elsewhere in Europe and Arnaut de Zwolle's manuscript.
Keyboards were usually made of quartered beech; maple was used in some south Italian
harpsichords. The end of the key-lever was guided by a wooden tongue in a vertical slot on the
rack. The travel of the keys was arrested by cloth padding on the front key-frame rail or by the jacks
reaching the padding on the jackrail, or probably sometimes by a combination of the two.
Although no unaltered action survives, the amount of sharp projecting above the natural-key
covers indicates a fairly shallow depth of touch (5–6 mm) in many instruments. The natural keys
were usually covered with boxwood, or with ivory in especially fine instruments; only rarely were
dark woods such as ebony used. Sharp keys were normally made of black-stained pear wood
topped with a thin slip of ebony.
Italian jacks were usually of a pear-like wood and about 5 mm thick, thicker than those used in
other countries, adding weight to compensate for their short length. Small springs of flat brass
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 2/9
strip were used rather than boar-bristle. The centrally-placed tongue enabled a damper slot to be
cut on both sides. Most instruments had quill plectra and a one-piece boxslide about 2·5 to 5 cm
deep. These were often made by gluing small blocks of wood to a thin strip, with the correct
spacing for the thickness of the jacks, and then gluing another strip on the open side. The jackrail
was usually decorated with the same mouldings as employed on the case. To hold it in place,
slotted blocks were glued to the inside of the case. Many harpsichords (mostly from Venice) had
the line of the jacks running not at 90° to the long side, but at such an angle that the jackslides
were nearer the front of the instrument at the treble end. Makers may have chosen this
arrangement because it reduces the amount of curve in the bentside (if other factors of scaling
and plucking points are unchanged).
Many 16th-century harpsichords have cypress soundboards, usually made of quartered timber,
but spruce and what appears to be fir were also used. Maple, whose mechanical characteristics
are similar to those of cypress, appears in some Neapolitan harpsichords. Cypress was
frequently used for the bridge when the soundboard was of cypress; walnut and beech were also
employed. The bridges were always parallel-sided with a moulding on the top edge; the crosssection
was normally smaller than in instruments from other countries. The height of the bridge
was usually reduced towards the treble. Double-pinning with a high hitch-pin rail was not used in
16th-century Italian harpsichords. Instead of the sawn or bent curve of the bridge in the bass
found in other countries, Italian harpsichords usually had a small piece mitred at an angle to the
main bridge for the last few notes. Nuts were usually of the same material and finished to the
same dimensions as the bridge, which has enabled the detection of many later alterations. They
were either fixed on a straight line, or in a curve with its inside facing the jacks. A curve in the
opposite direction results from later shortening of the strings. With the scales and plucking points
chosen by Italian makers, the two nuts in a 1 × 8′, 1 × 4′ disposition lay quite close to each other,
making it impractical to locate the 4′ tuning pins between the 4′ and 8′ nuts (the commonest
practice outside Italy). Instead, holes were drilled through the 8′ nut so that the 4′ strings could
reach their tuning pins at the edge of the wrestplank. One of the few harpsichords of this type to
have retained its original 8′ nut is by Francesco Padovano (1561; Deutsches Museum, Munich).
The 4′ hitch-pins were sometimes simply driven into the soundboard and secured with a drop of
glue. This practice is only possible with a relatively dense wood such as cypress (not with spruce
or fir), but thin 4′ hitch-pin rails glued to the soundboard were also used. The 8′ strings were
hitched to the soundboard liner in the conventional way.
The nut was placed on (or near) solid timber in all known 16th-century Italian harpsichords and
does not contribute to the audible vibrations. Some earlier sources incorrectly interpreted the
cypress veneer (c6 mm thick) that is often found on wrestplanks as being an additional
soundboard.
Ribbing systems have been found with three or four crossbars running at an angle from the spine
towards the front of the instrument and crossing under the bridge, where they are usually
undercut to leave the soundboard free. Others have a cut-off bar, with or without additional
crossbars. Some harpsichords seem to have been made without any bars at all. The
impossibility of access to the inside of many instruments makes it difficult to establish how rigidly
makers followed these systems; exceptions can be found. These barring systems are found in
Italian harpsichords from the 16th century to the 18th; no feature can be categorically assigned to
one period, and no specific conclusions can be drawn about the sound of a harpsichord simply
from the type of barring used.
The point at which a string is plucked is important in determining the character of the instrument's
sound. When the plucking point is near the nut (close plucking) the sound is nasal; nearer the
middle of the string (centre plucking) it is rounder. In Italian harpsichords of all periods the
plucking point of the back 8′ register (furthest from the player) lay at close to a third of the string
length at c″. At the extreme treble the plucking point was nearer the middle of the string. In the
bass the plucking point was, in order to avoid over-long key lengths, relatively close to the nut.
Italian harpsichords with a 1 × 8′, 1 × 4′ disposition had the 8′ in the back register with the jacks
plucking to the left. 21 examples of this disposition are known from before 1600. A comparison
between this arrangement and that of Ruckers's harpsichords (see §3(i) below) reveals a basic
difference of design, and hence of sound: Ruckers harpsichords have the 8′ plucking to the right
and in the front register, giving a more nasal sound. The Italian harpsichord is a little sweeter, and
in 1 × 8′ instruments the 8′ register was generally in the same position and had the same
plucking point as in the disposition with 8′ and 4′.
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 3/9
Italian harpsichords are typically described as having a bold sound with a more pronounced
attack than in other harpsichords, but this judgement has chiefly arisen from listening to brassstrung
17th- or 18th-century harpsichords. Since practically every 16th-century Italian harpsichord
has been modified in some way that affects tone, even the few playable examples are not a
reliable guide to how these instruments would originally have sounded. Moreover, most 16thcentury
harpsichords were intended for iron stringing (see below); iron-strung instruments tend to
have a more brilliant sound with a longer decay time than brass-strung ones. Because the
majority of the harpsichords that survive from before 1600 were made in Venice, with only a few
from elsewhere in Italy, it is not possible to generalize about regional differences in harpsichord
tone.
Although a number of 16th-century harpsichords now have two 8′ registers, many of these have
had a register added to what was originally a 1 × 8′ disposition. An example is the harpsichord of
1521 by Jerome of Bologna (Victoria and Albert Museum, London; see HIERONYMUS
BONONIENSIS), previously thought to be the oldest surviving harpsichord. This status is now held
by an instrument inscribed as being started on 18 September 1515 by Vincentius. It too probably
had a single 8′ register, with a compass of FGA–g‴a‴ or perhaps C/E–f‴. Since the majority of 1
× 8′, 1 × 4′; harpsichords were built in Venice the 4′ stop might appear to be a Venetian invention,
but the paucity of evidence from other towns imposes caution. In any case, given the prevalence of
15th-century instruments at 4′ pitch it may be more accurate to say that an 8′ stop was added.
Many 16th-century Venetian harpsichords had their 1 × 8′, 1 × 4′ dispositions altered to 2 × 8′ after
about 1630 (see §3(iv) below). The 1574 Baffo is one such instrument. Later scholarship
recognized that of 50 harpsichords known from before 1600 only eight were made with two 8′
registers and nine with a single 8′ register. The earliest dated 2 × 8′ specification was built by
Domenico da Pesaro in 1570. It is likely that the four ‘gravicembali doppi’ used in Florence in
1565 at the wedding celebrations for Francesco I de' Medici were 2 × 8′ harpsichords. The earliest
known 2 × 8′ harpsichord (signed ‘Bortolus’) was made probably in the 1540s for the court of
Ercole II d'Este, Duke of Ferrara. An unusual type of 2 × 8′ harpsichord had the jacks facing each
other on either side of a narrow-spaced (3 mm) pair of strings (e.g. Celestini, 1569; Royal Ontario
Museum, Toronto). This required an unsual S-shaped end to the key levers. Both registers had
strings of almost identical length; the system may have been intended to improve tuning stability
(Wraight, H1993; Lee, A1996 and A1997). Although many instruments were built in the 15th
century at 4′ pitch, only two 16th-century octave harpsichords are known, both made by Domenico
da Pesaro (mentioned above).
A discussion of compasses must take into account the alterations that obscure the original
condition of many instruments, first noted by Barnes (in Ripin, A1971). Only one of the known
16th-century Italian harpsichords has not had its compass, disposition or scale altered (Wraight,
H1997). The compasses described here as the original ones are mostly not the present ones.
Around 1500, harpsichord compasses probably still began with FGA, that is, lacking F and G .
These compasses may have reached as high as f‴, as in the Urbino intarsia clavichord of
around 1476 (see CLAVICHORD, [not available online]), or only to a″ (probably without g ″) or c‴. An
intarsia of a virginal (probably made in 1496 by Lorenzo da Pavia) in the grotta of Isabella d'Este's
study in Mantua shows a compass of C/E–c‴, which could also have been used for harpsichords
at this time. In the 16th century the most common compass for harpsichords or virginals was
C/E–f‴. The C/E–c‴ compass was used in only a third of surviving harpsichords. An early
harpsichord with an exceptionally wide range and low pitch is the 1579 Baffo (Musée de la
Musique, Paris), which originally had a compass of C/E–c′′′′, although the sounding range was
G′–g‴. Chromatic bass octaves were apparently not used before 1600 and were rare thereafter. It
is unlikely that harpsichords were made with a compass of G′/B′–c‴ before 1600, although
several instruments, previously C/E–f‴, were later modified to this range (e.g. the 1574 Baffo in
fig.4 below). A compass of G′/A′–c‴ was known from the 1630s and was common towards the
end of the 17th century, but probably was not used in the 16th century. The inventory dated 1700 of
Medici instruments lists a harpsichord of 1538 made by Domenico da Pesaro with a 50-note
compass of G′/A′–a″, but this may have been the result of an alteration of a 50-note C/E–f‴
compass.
Temperaments of the 15th century to the 17th (see TEMPERAMENTS, §§1–5) often gave chromatic
notes that were not enharmonically equivalent (e.g. G or A , which were not at the same pitch); to
provide keyboards with the missing notes, extra chromatic keys (usually D and A ) were
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 4/9
sometimes added. This practice seems to have originated in Italy, where organs were furnished
with at least one split key in some octaves as early as 1468 (Wraight and Stembridge, H1994).
Although there was interest in this approach in other countries, such keyboards are found mostly
in Italian harpsichords and virginals beginning about 1620. The earliest surviving harpsichord
made with split keys for D /E and G /A was built for the court of Alfonso II d'Este at Ferrara
around 1570 (no.1883.718, Kunstgewerbemuseum, Schloss Köpenick, Berlin). For discussion of
the most important experiments with enharmonic keyboards, see ENHARMONIC KEYBOARD.
Detailed studies to determine the original pitches of Italian harpsichords began in the 1960s.
Ascertaining their pitch is essential to understanding their musical function. Although many
instruments were at 8′ pitch or its octave, some others were a fourth higher or lower than 8′; the
purpose of such instruments and how they combined with other instruments is still a matter for
study. Pitch is related to string length and also whether the instrument is strung with brass or iron
wire: the string material imparts a specific timbre to the tone. The alterations that have been made
to surviving instruments have tended to complicate discussions about pitch. Italian string lengths
in virginals and harpsichords ranged from about 15 cm to 42 cm measured at c″ (the short end of
this range mostly being small virginals), but the usual range for 8′ instruments was about 25·5 cm
to 36 cm. Thomas and Rhodes (1967) suggested that iron wire, which permits a higher pitch,
was used for instruments with longer scales and brass for those with shorter ones; Barnes
(1968; in Ripin, A1971; Barnes, H1971) argued that brass wire was used for all instruments and
that pitches varied among instruments in proportion with string lengths. Later scholarship
provides better data about the original scales of many instruments. It might at first appear that the
wide range of string lengths among instruments allowed for considerable latitude of pitch,
particularly because it is possible to tune a string over a range of pitches below its breaking point
and still produce an acceptable tone. Wraight's work (Early Keyboard Journal, H2000), however,
suggests that among 16th-century instrument makers in Venice (where the majority of surviving
instruments were made) a range of 8′ pitches (a′ = c440–490) was in general use. It appears that
these makers regularly and accurately used the same scales, with closely defined string lengths;
there was agreement on this not only within individual workshops but also among different
makers. Some later modifications to 16th-century instruments show that makers considered it
desirable to alter the scale of an instrument even when changing its pitch by only a semitone.
There is clear evidence that 18th-century makers such as Cristofori, Ferrini and Solfanelli used
both iron and brass wire to string some of their instruments, and that a ratio of nearly 5:6 for the
lengths of brass wire and iron wire at the same pitch was consistently employed (O'Brien, A1981,
and Wraight, H1997). The range of scales found before 1600 would seem also to allow for the
use of either brass or iron wire, and documentary sources establish that both were used; the
problem is to identify the stringing material for each individual instrument. In general, 16th-century
Italian instrument makers seem to have preferred iron strings regardless of instrument type, size
or compass; in any case, the stringing material of chamber keyboard instruments in this period
was not exclusively linked to compass or scale. Most 16th-century harpsichords originally had c″
at about 30 to 35 cm. Many of these instruments had a 4′ stop and a compass of C/E–f‴;
available evidence indicates that both harpsichords and virginals with scales of this length were
intended for iron strings. Galilei, in his Dialogo della musica antica e della moderna (1581),
suggested that the ‘gravicembalo’ had iron strings in the treble and brass in the bass, although
he did not specify how far into the bass the iron stringing extended. The scale design of these
harpsichords would require brass wire only for the last few notes and implies that iron-strung,
long-scaled harpsichords would have stood within the normal 8′ pitch range. (The name
gravicembalo does not, as it might appear, indicate a low-pitched instrument at this period; it
might originally – around 1500 – have meant a harpsichord at 8′ pitch as compared to the
prevailing 4′ pitch of chamber keyboard instruments of the time.)
A few harpsichords have a scale with c″ at about 30 cm, but are without a 4′ stop and have
compasses that do not reach to f‴. The Italian tradition of scale design indicates that these were
also intended for iron strings; like virginals with the same scale length, they were pitched a tone
above those instruments where c″ is at 33 cm. Some harpsichords with very long scales, c″ being
at 41 to 47 cm (e.g. the instruments by Baffo, 1574 and 1579, and Francesco Padovano, 1561,
mentioned above), would have been pitched a 4th lower than those with c″ at 30 to 35 cm, even if
strung with iron wire. Harpsichords with short scales, where c″ is at 27 to 29 cm, might at first
appear to be intended for brass wire at normal 8′ pitch. Wraight's analysis of the scale design,
including the bass strings, implies that all such instruments were probably intended for the
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 5/9
higher 8′ pitch (a′ = c520) with iron wire. Examples are the 1521 Hieronymus Bononiensis
instrument (discussed above) and the ‘Rigunni’, probably of 1584 (Stearns Collection, Ann Arbor,
Michigan). There seem to be a few harpsichords that were pitched a 4th above 8′ pitch if iron
stringing is assumed (e.g. Celestini, 1608, Museum für Kunst und Gewerbe, Hamburg; Celestini,
1596, Royal Ontario Museum, Toronto). Octave pitch, for example in the harpsichords of 1543 and
1546 made by Domenico da Pesaro, was the highest pitch in normal use in the 16th century.
The coordination and standardization in scale length among Renaissance harpsichord makers in
Venice need not imply that performance pitch was as well organized; there were enough
intermediate sizes of instrument that pitch incompatibilities in performance could easily have
arisen. A striking feature of many Renaissance Italian compasses is that they end on either c‴ or
f‴. Some scholars of the 1960s linked this with scaling, suggesting that harpsichords and
virginals ending at f‴ tended to have longer scales and were pitched a 4th lower than those
ending at c‴. It is now clear, however, that most of these instruments sounded at 8′ pitch and that
the compass ending on f‴ simply reflects the Italian tradition of extending the musical range only
in steps of a 4th (or 5th) and much less frequently by individual notes (e.g. from c‴ to d‴). The
high f‴, sounding a high pitch, would have facilitated the performance of music at octave pitch
(printed music rarely went into this range).
Denzil Wraight
(ii) Northern Europe.
15th-century documentary evidence suggests that string keyboard instruments were first
developed in northern Europe. The oldest surviving plucked string keyboard instrument, a
CLAVICYTHERIUM of about 1480 (Royal College of Music, London), was made in Germany. Most
surviving 16th-century string keyboard instruments, however, are Italian; and the earliest known
from northern Europe, a harpsichord made by Hans Müller in Leipzig in 1537 (Museo degli
Strumenti Musicali, Rome) and a virginal made by Joes Karest in Antwerp in 1548 (Instruments
Museum, Brussels Conservatory; for illustration see VIRGINAL, [not available online]), share many
characteristics of Italian instruments made decades earlier, such as thin case sides surrounded
by applied mouldings. It was thought (Ripin, A1971) that the style of Müller and Karest was
derived from Italian models, but it now seems more likely that 16th-century Italian harpsichordmaking
traditions had origins in 15th-century north European practices. The German
clavicytherium of about 1480, which except for its upright form probably resembles a normal
harpsichord of the period, anticipates several characteristics of 16th-century Italian harpsichords:
thin case sides attached to the edges of the bottom board (the back in the clavicytherium),
scrolled cheeks, a very acute angle at the tail and a separate outer case. The clavicytherium also
resembles the clavisimbalum described by Arnaut de Zwolle about 1440: both had nonPythagorean
scaling (see §1 above), a relatively shallow space (5 cm in the clavicytherium)
between the soundboard and the bottom board, and multiple roses in the soundboard. The lower
guide in the instruments of Müller and Karest, consisting of a thin plate of wood covering the
entire area over the portion of the keyboard behind the nameboard, may be a vestige of the
clavisimbalum’s bottom board (which was placed above the keyboard in one of Arnaut’s
designs), while Müller’s key-guiding system, with the distal end of the key lever forked for a vertical
pin held by the back rail, is the same as that in the clavicytherium. Karest’s instruments have
multiple roses and use the proportions that Arnaut specified for his clavichord. The simplest
explanation for these and other correspondences is that 16th-century traditions in both Italy and
northern Europe were separate branches of an earlier northern tradition. This does not, of course,
preclude the possibility of subsequent Italian influences on Northern practice.
Of a small number of surviving 16th-century keyboard instruments from northern Europe, about 20
are virginals; only two securely dated before 1590 are harpsichords. Documentary sources are
scant, and north European depictions of harpsichords are rare compared to representations of
clavichords and virginals. It is evident, however, that the major traditions of north European
harpsichord making became firmly established during the 16th century, although knowledge
about such details as string scaling and case construction must be derived primarily from
virginals. The Müller harpsichord of 1537 was made in a style distinct from that of Italy. The bottom
board is only 8 mm thick; it is attached to the lower edges of the sides, which are 7 to 8 mm thick
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 6/9
and of softwood veneered with Hungarian ash. The soundboard extends to the nameboard and
has a separate mortise to guide each jack. Because the wrest plank is only about 5 cm wide, the
nut is on active soundboard. Some details of the original disposition and scaling are obscured by
later alterations. It was certainly made with two sets of strings and three registers, one of them a
nasal (lute) stop very close to the nut. The original stop-changing mechanism (consisting
probably of movable lower registers placed over the oversized mortises in the lower guide) is
missing, but there are holes for knobs to project through the cheek, including one probably for an
ARPICHORDUM STOP. The keyboard, originally CD–g″a″, could be shifted to change the sounding
pitch by a whole tone (see TRANSPOSING KEYBOARD). There appear to have been two bridges and
two sets of strings a 4th apart (Koster, F1996). Müller’s scaling is foreshortened in the bass,
more like that of 15th-century instruments than the typical Italian harpsichord scaling, which is
Pythagorean almost to the lowest note. The foreshortened scaling may imply the use of iron
strings in the treble and brass in the bass; it is also related to the reverse curve of the bridge in
the bass and the straightness of the bentside from its midpoint to the tail. The ungainliness of the
instrument’s outline in comparison with Italian harpsichords, however, arises largely from the
added width (about 7·5 cm) required by the shifting keyboard and the short length associated with
scaling for a high pitch.
Like the Müller harpsichord, the two surviving virginals made by Karest (1548 and 1550) have
moderately thin case sides outlined with applied mouldings, a plate-like lower guide and
foreshortened scaling. Although their absolute pitch levels are disputable, the instrument of 1550
is the larger and was probably tuned a 4th lower than the other (the f″ string in the larger
instrument being roughly the same length as the c″ in the smaller). Karest and Müller may have
applied the putative archaic principle that string lengths, at least in the treble, should be
equivalent to the speaking lengths of organ pipes of the same pitch; that is, they may have used
low-stress iron scalings, so that an instrument tuned to normal 8′ pitch would have a c″ string of
about 32 cm. Karest’s 1548 virginal would thus have sounded approximately a semitone above
modern pitch. Some later German and Austrian instruments, stylistically similar to Müller’s and
Karest’s, used such scalings (Kukelka, F1994).
Except for instruments made in the Low Countries after about 1575, most north European
harpsichord making before the 18th century has stylistic affinities with Müller’s and Karest’s
works. There is documentary evidence that instruments by Karest and other Antwerp makers
were sent to Germany during the 16th century, which may have transmitted some of their
traditions; or Müller’s and Karest’s instruments may simply represent a style that emerged
gradually throughout northern Europe, spread partly perhaps by organ builders, who were
necessarily itinerant and who also made string keyboard instruments. The term ‘international
style’ has been applied to this group of tendencies and techniques, which include relatively thin
case sides, plate-like lower guides, nuts placed on resonant soundboard, light 4′ hitch-pin rails,
foreshortened scalings and provision for a variety of tone colours. Although most surviving
harpsichords made in this tradition, which extended from France to Sweden and from England to
Austria, date from after 1600, a widespread inclination to make complex instruments is evident in
early inventories. The 1566 Fugger inventory (see Smith, C1980), for example, includes an
English harpsichord with several registers, an instrument made in Cologne with two keyboards
for two performers, and one from the Netherlands with four keyboards for four performers.
Virginals with thick case sides and long iron-string scaling in the upper register (c″ = c36–38 cm)
made in Antwerp about 1580 by Hans Bos, Hans Ruckers and others show that the basic style
practised by the Ruckers family and other Antwerp makers throughout the next century was
already well established. Together with an anonymous virginal dated 1568 (in the Victoria and
Albert Museum, London), these instruments show the development of the layout and internal
construction of muselars and spinetten. Unfortunately, no well-preserved Antwerp harpsichords
survive from this period, when harpsichord making presumably underwent analogous
developments. Some idea of a transitional style of the 1560s, however, is provided by a
harpsichord made in London in 1579 by Lodewijk Theeus (ii), who became a member of the
Antwerp Guild in 1561 but had emigrated to London by 1568. Although some of its features, such
as the use of oak for the case, the chromatic compass in the bass and perhaps the 2 × 8′, 1 × 4′
disposition, may be regarded as English, others presumably reflect the Antwerp style of the mid-
1560s (Koster, D1980). As in the Müller harpsichord, the rear portion of the bentside is straight
and the soundboard, mortised to serve as an upper guide for the jacks, extends to the
nameboard, so that the nuts are on active soundboard. Because the wrest plank is narrow, the 4′
wrest pins are grouped with the 8′ pins, and the 4′ strings pass through holes in the 8′ nut. The 4′
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 7/9
hitch-pin rail is exceptionally light in comparison with those of later Antwerp harpsichords. The
case sides are about 13·5 mm thick, and the lid is hinged to the spine. Mouldings applied to the
interior of the walls give the illusion of a thin-cased inner instrument inside a massive outer case.
The scaling, with the longer of the two c″ strings about 35·6 cm, is foreshortened in the bass. The
pitch, estimated from a pipe in the organ with which the harpsichord was combined shortly after it
was made, is about a semitone below modern pitch. While the disposition of Theeus’s
harpsichord is different in detail from that of Müller’s, the intention of both makers must have been
to provide a wealth of tone colour. Since Theeus probably placed his 4′ jacks in the central
register, the 8′ stops, with widely separated plucking points, would have been quite different in
timbre. Stop knobs at the front of the instrument moved the lower guides to change the
registration. (Movable lower guides, with the soundboard used as a stationary upper guide, were
apparently also used by Müller and may have been a common north European characteristic.) A
set of large bent pins in the bridge was evidently intended as a permanently engaged
arpichordum stop for the shorter set of 8′ strings. Although the later standard Antwerp harpsichord
disposition (1 × 8′, 1 × 4′) is decidedly less colouristic than the dispositions of Müller and Theeus,
a relish for varied timbres is still evident in the development of muselars, spinetten and motherand-child
virginals, which can be coupled together to provide an 8′ plus a 4′ registration. In view of
these it seems possible that harpsichords with complex dispositions were made in Antwerp
before Ruckers. The earliest extant Antwerp harpsichords, made about 1590, are, however, nearly
identical to the standard Ruckers instruments of the 17th century in disposition and layout. They
differ structurally from the Theeus harpsichord, most importantly in that the nuts are placed on a
solid wrest plank and the provision of a much heavier 4′ hitch-pin rail.
During the 16th century north European makers began to build harpsichords that were wider to
allow larger keyboard compasses, and longer to accommodate longer, lower-pitched strings. The
presumably typical compass described by Sebastian Virdung in 1511 was FG–g″. As late as the
1570s some instruments were still being made with FGA–g″a″, commonly used in organs. C/E–g
″a″ had become customary in Antwerp by the 1540s and is found even in some instruments made
there in the 1590s. Karest’s 1548 virginal, however, already had C/E–c‴ which remained usual
on the Continent well into the 17th century. The Theeus harpsichord (C–c‴), as well as an
English depiction of a virginal dating from 1591 and the use of low accidentals in English
keyboard music of the late 16th century, all indicate that the chromatic bass octave (sometimes
lacking C or perhaps with the apparent C key tuned to A′) was a characteristic feature of
English harpsichords.
The German clavicytherium of around 1480 was probably tuned about a 4th above 8′ pitch, and
the Müller harpsichord of 1537, even at the lowest level afforded by its transposing devices, was
undoubtedly designed for a high pitch. The Karest virginal of 1550, however, could not possibly
have been tuned higher than 8′ pitch and may have been significantly lower. By the end of the
century, harpsichords with two keyboards, one at 8′ pitch, the other a 4th lower, had been
developed in Antwerp. The earliest dated survival (Händel-Haus, Halle) was made in the Ruckers
workshop in 1599, but two anonymous examples (Instruments Museum, Brussels Conservatory)
may date from the 1580s. In all three instruments, before later alterations, the low-pitch keyboard
had a compass of C/E–d‴. Instruments at high pitch continued to be made; the tradition of
making instruments at various high and low pitches, seen most systematically in the work of the
Ruckers family, persisted through the mid-17th century.
Some 16th-century German inventories hint at the existence of harpsichords with two manuals,
although the generic term ‘instrument’ might refer to mother-and-child virginals or rectangular
instruments with keyboards for two players at different sides of the case. Even if the instruments
were wing-shaped harpsichords, the two keyboards may have been at different pitches, as in the
transposing doubles made in Antwerp. An ‘instrument with two ivory keyboards, purchased in
Frankfurt an der Oder’ listed in a Dresden court inventory of 1593 (transcribed in Fürstenau,
C1872), however, have been a true non-transposing two-manual harpsichord, since mother-andchild
instruments are described explicitly in the same source. Given the evident north European
fondness for contrasting tone colours in harpsichords and the model provided by organs with
multiple keyboards, it would be remarkable if non-transposing two-manual harpsichords had
never been made during this period.
John Koster
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 8/9
Copyright © Oxford University Press 2007 — 2011.
(iii) Spain.
The harpsichord does not appear in Spanish iconography until the late 15th century but
documentary references date back to the mid-15th. Juce Albariel, known as the Moor of Zaragoza,
was described in 1465 as a maker of lutes, clavichords and instruments. He may have been
responsible for a clavicimbalo in a black case inventoried in Zaragoza in 1469. Zaragoza was a
notable centre of keyboard instrument making in the late 15th century and the early 16th, its most
famous representative being another Moor, Mahoma (Joan) Mofferiz, who made instruments for
royalty and the nobility, including a claviorgan with both gut- and wire-strung registers. In Seville,
Maestro Enrique was building clavicimbanos in 1470, while before 1502 the Sevillian carpenters'
guild required apprentice luthiers to learn how to make a clavizimbano.
Evidence for harpsichord making in the 16th century is entirely documentary. The richest source is
the inventory of instruments belonging to Philip II (reigned 1556–98). His largest clavicordio (i.e. a
plucked keyboard instrument; see Kenyon de Pascual, I1992) was about 223 cm long and the
smallest (a triangular instrument) only about 42 cm. Ebony and maple are the woods most
frequently mentioned in the inventory, although a small instrument (c55·5 cm long) made by the
Moor of Zaragoza was of inlaid walnut. Many of the king's instruments may not have been made by
Spaniards. Following the installation of the Habsburg dynasty on the Spanish throne with the
succession of Charles I (Charles V) in 1516, instruments and instrument builders were brought
to Spain from the southern Low Countries, while there were close ties with southern Italy and
Milan, which were Spanish possessions. One might, therefore, expect some Spanish
harpsichords to have shown features found in Flemish and Italian instruments.
Beryl Kenyon de Pascual
14.3.2011 Harpsichord, §2: The Renaissance in …
…kb.nl/subscriber/article/…/12420pg2… 9/9