Bent Plates in Violin Construction

Author(s): Athanas Lolov

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

Galpin Society is collaborating with JSTOR to digitize, preserve and extend access to The Galpin Society Journal.

http://www.jstor.org

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions
 ATHANAS LOLOV

Bent Plates in Violin

Construction

T HE use of bent top plates in the construction of viols was recently

discussed by Dietrich Kessler.1 He presented convincing evidence

that a viol top by Richard Meares (c. 1660) was formed at least partially of

slats bent in the manner of a lute back, and proposed that some method

of bending might have been used by other well-known English makers,

most importantly Barak Norman. The discovery of what may have

been a school of bent-plate viol builders in England is important in

itself, but is particularly interesting for the support it offers another,

even more controversial, theory: that bending could have been used by

some of the Cremona violin makers.

In his article concerning the Meares viol, Kessler rejected any

application of the bending method to two-piece tops such as are used for

violins, saying only that the bending method does not apply to violins

because 'no bending is used in their construction'.2 No support was

offered for that statement, it being simply an assumption, and that is a

little surprising given the innovative nature of Kessler's article. Most

people had assumed no bending to be involved in viol tops either, until

he suggested it. The present study deals with a violin belly which was

bent from slats Y56" thick using a method very different from that

employed by Kessler; the resulting plate has characteristics that could

suggest an answer to some of the questions regarding the peculiarities of

violins built by the masters of Cremona.

While carving from a thick plank is certainly the most familiar

method for making a violin belly, and one with a long history, we

actually have no hard evidence that Stradivari's tops, for example, were

'dug from a plank'. We just have assumed that they were. A respect for

the scientific method requires at least that this admittedly somewhat

outlandish possibility be investigated. Indeed, it is curious that in all the

history of the Stradivari mystery, during which innumerable eccentric

or preposterous ideas have come forward,3 no one, to my knowledge,

has ever proposed that the master's tops might have received their initial

form by bending. That the idea is not so unnatural is evident from the

way I first heard it: a musically naive visitor looked at a newly

completed viol and asked, 'how do you bend the tops ?' What seemed at

10

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions
 . .... . . . :i::;::::j:~_:-_--

ON,

IN
_:- : _: -. :: :: :: ::: .:: _-_ . :-i: _-.-ii-ilii~ii- iii-i~ -i- ii- _?::-- I_ -:- :::::-::,: -:: ::I .:- .:.-::: _::_ :-::i:?:: ::- ::: :-- :::::~l~iiiiiiiiiiiiiiiiizl~ :.:: : ': :......... ...........:: .. : :
:::.i~i-.ii ii~ii:::: ................ .; -

......... .

,. . . . . . . . . . . .

RE

IN

4.1?1 . . . . . . . . . .

./............ ..... . . . . . . .

(b)M

PLATE IV

Violin top made from bent plates. (a) before and (b) after joining.

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions
first like a humorous error led to the building in 1981 of a large viol, on a

Barak Norman pattern, with a top formed by bending. I found it to be

an excellent instrument, powerful in tone, quick to speak, and

remarkably free of wolves. The bending method I used was one which

could be applied to a two-piece top, and it was not long before I began

to wonder about the application of bending to violins.

The violins of Stradivari and those of the Guarneri family are alike in

that they are regarded as possessing a mysteriously irreproducable

beauty and power of tone. They also share some physical peculiarities

which are not easy to explain in the context of the traditional method of

building:

1) The ribs usually taper from the side blocks to the top. Neither the

amount of tapering nor the maximum height are consistent from violin

to violin, and the usual explanation for this tapering, that the belly thus

offers 'the necessary resistance to the neck',4 is questionable. William

Hill said of this feature, 'We entirely fail to perceive any fixed idea or

principle guiding Stradivari in determining the relative height of

sides . . ..

2) The f-holes almost always show distortion, the lower wing being

sunk beneath the plane of the plate. This effect is, of course, most

pronounced on the right side, where the sound post helps to support the

belly, but it does not appear in violins of later eras except when induced

by artificial means. Edward Heron-Allen, in his classic work on violin

construction, recommended producing this effect by means of twisting

the wood with a small stick, a procedure which smacks of superficial

imitation.6

3) The grain of the top often appears to converge slightly at the

extremities of the center joint. This phenomenon is largely hidden by

the fingerboard and tailpiece, but is still visible in vertical photographs

of many master instruments.7 The convergence is most easily observed

in instruments with very high arching. As long as the grain at the outer

extremes of the instrument remains straight, the curving of grain toward

the center could be a sign that the wood was bent upward in the middle

(see Fig. 1 for a graphic representation). The effect is, to be sure, very

subtle, and easily confused by irregular grain, but in those instruments

where it is plainly visible it is hard to explain by any other

hypothesis.

Some of the historical evidence also poses questions:

1) Stradivari is reported to have preferred to split his own wood for

tops. This implies that he did not use wood seasoned for many years.

Could he have been moved by any reasons other than economy? Could

a thick plate dry sufficiently in a short period to be safely employed in

11

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions
 a violin? If he cut relatively thin slabs from green wood and wedged

them immediately, they would dry sufficiently in a single season and

retain their curve without further treatment.

2) It has often been suggested, as by Edward Heron-Allen in the

nineteenth century, that master builders split wood for violin tops,

rather than sawing it, to avoid disturbing the longitudinal fibers.8 Yet

how can one avoid disturbing the fibers when carving begins ? If the

grain follows the arch rather than being cut to it, more of the

longitudinal fibers will remain whole, and the top will be a great deal

strengthened.

The example shown in Pl. IV b was bent by a simple process. The

wood used was two book-matched slabs of Sitka spruce, each /16" thick.

They were thoroughly soaked in water, and then clamped together

along three sides. (It is worth noting that the soaking in water could have

been skipped if freshly-cut green wood were available.) A pair of

wedges was then driven into the areas corresponding to the upper and

lower bouts, the tips of the wedges being directed toward the center of

the curve of each bout. There are choices possible here; both the

orientation and the shape of the wedges could be altered with possibly

significant results. The wedging easily produced a recognizable

approximation of the flat-topped curve characteristic of Stradivari's

violins.

The wood, being thin, dried quickly, and retained its new shape well.

Beforejoining, the bent slabs presented the appearance of having grown

apart at the center of their joint (see Plate IV a), a consequence of

bending which necessitated spot-gluing each to a plank in order to plane

the joint straight. This is a step requiring some skill and practice, owing

to the curving of grain introduced by the bend. Once glued, the grain at

the tail end of the joint can be seen to converge slightly, though

irregularity in the grain has made this effect invisible at the upper

end.

The plate does not lie flat; the edges are warped upwards in the

middle as a consequence of stresses induced by the bending process. As

with the viol plates described by Kessler, some pressure is required to

force it down to meet the ribs. This provides a plausible motive for

tapering the ribs slightly, as less pressure would then be required to

make the joint. Even with this partial match of ribs to belly, however,

some pressure is still required; this pressure causes a small distortion of

the wings of the sound holes, resembling that which Heron-Allen

imitated by use of a stick.

The amount of distortion in the plate is unpredictable, being

different for each piece of wood. This could explain not only the

12

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions
 variations in the distortion of f-holes, but also the seemingly random

variation in the rib height of Stradivari's works, assuming an attempt

was made to shape the ribs to at least partially conform to the top.

Carving was minimal, and was guided by the roughly defined shape

already present in the wood. An added benefit was the ability to follow

the fibers with the plane: because the grain was rarely interrupted by the

face of a curve, the plane glided easily, even travelling 'uphill' without

digging in. The relative lack of exposed fiber ends should also make

finishing easier.

It is clear that the labor saved by this method is considerable, but that

might not have been the only motive for a maker to bend tops. In the

seventeenth century the cost of materials made a much more significant

proportion of the cost of any manufactured item, and that could have

caused a craftsman to consider ways by which he might double the

number of instruments obtainable from a given amount of wood.

Whether the method was adopted from economic motives or from an

understanding of the benefits in structural strength is, however, of little

importance. If it was indeed used by the Cremona makers it was used

because it worked, and perhaps was kept as a trade secret for the same

reason, a feat which would not be difficult since the bulk of the

labor in violin manufacture was concerned with the other parts

of the instrument. Kessler's discovery of a (probably not unique)

viol belly made by bending in seventeenth-century England is proof that

even such a large trade secret as this one can be both kept and

forgotten.

Further research should focus on an examination of the center joints

of a wide range of master violins, and, if possible, on a microscopic

examination of the fibers in a section of the arch, should a fragment of

one of these instruments ever be available. Convergence of grain at the

top and bottom of the joint is an unavoidable artifact of the bending

process but is not always apparent to the eye, as is evidenced by the plate

described above. Yet, even in this plate careful use of a straightedge

with the photograph will reveal a slight convexity of grain on both sides

of the joint; instruments displaying this characteristic may justly be

suspected of having been bent. Fig. 1 provides a clearer image of the

effect of bending on grain. The drawing was made by photographing

ruled paper bent to roughly the curve of a violin belly; the lack of

irregularity in the 'grain' makes the convergence at the center joint

more obvious, though if the center joint were not visible the curve of

even these ruled lines would not be strikingly apparent. Stradivari's last

instruments were somewhat flatter than this model and any curving of

grain will consequently be still more subtle.

13

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions
 __

__?_ ??_ _?____

?_ ?_
?_

_?? ____ __~___

_ ?_? ?

__? ? ~
__ ? ?
I _I _I __

I_ ? ~
__

_??1
1

___

________

_ .. __ ___~_-

--

__

FIG. 1. Simulation of bent plate top.

It is unfortunate that close examination of a master instrument is very

difficult to arrange, as a much higher degree of certainty could be

obtained merely by looking under the tailpiece and fingerboard with an

illuminated mirror. The superficial examinations already made, without

the opportunity to look beneath tailpieces and fingerboards, indicate

that many of the violins thought to be from Cremonese masters have

signs consistent with what would be expected from a bent top, while

instruments from other schools and later eras generally do not. This

suggests an interesting possibility, that if the bent-plate theory is proved

a fact, the violin trade will be provided with a powerful tool with which

to judge the authenticity of supposed Cremonese violins. 'Le Messie',

the most famous of the violins attributed to Stradivari, for example, is

alone among the instruments examined superficially in that it displays

none of the characteristics described above. A careful reading of the

story of 'Le Messie' shows that its history before its possession by

Vuillaume rests entirely on the veracity of that maker, who was famed

for his accurate copies of master instruments. Is it possible that 'Le

Messie' is one of them ?

There are many factors involved in the making of a fine violin, and

the manner of shaping the top cannot, by itself, necessarily be decisive

for the outcome. However, the bent-top theory is the first to provide a

plausible explanation for the unbroken string of failures which marks

the record of modern attempts to reproduce with consistency the tonal

qualities of Cremona violins. The varnishes have been examined with

x-ray spectroscopy; every dimension has been measured with all the

exactitude of which our technology is capable; the woods have been

matched so closely that variation between any two violins of one of the

old masters far exceeds the difference between a modern copy and its

original. And yet all the master violins, regardless of their variation in

14

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions
 shape, dimensions, and finish, share tonal qualities which set them far

above any of the copies. Even age cannot wholly account for the

difference, since the violins of Stradivari were famous even when he

was alive. Conversely, Villaume's copies are now close to two hundred

years old, and show no signs of beginning to sound like their models. A

basic structural difference between the fiber arrangement of the old

violins and that of the new would provide a rational explanation of the

phenomenon, though it would deprive us of the romantic hope,

cherished by many, that Stradivari and his peers were simply possessed

of magical powers, an idea which certainly would have amused those

superb craftsmen.

The author thanks Jeffrey Evans for his many suggestions during the

preparation of this paper.

NOTES

iDietrich Kessler, 'Viol Construction in 17th-century England', Early Music

10, No. 3 (July 1982), 340-345.

2 ibid., p. 345.

3 For example, it has been suggested that Stradivari's varnish contained

powdered diamonds, and that the wood for his instruments was cured by being

soaked six months in urine.

4See August Riechers, 'The Violin and Its Construction', in Violin

Iconography of Antonio Stradivari, 1644-1737, by Herbert K. Goodkind

(Larchmont, New York: Privately printed, 1972), p. 128.

5 William Hill, Antonio Stradivari, His Life and Work 1644-1737 (New York:

Dover, 1963), p. 55.

6 Edward Heron-Allen, Violin-Making as It Was and Is, 2nd Ed. (New York:

Carl Fischer, 1885), p. 258, fig. 155.

7 For examples, see Karel Jalovec, Italian Violin Makers (London: Paul

Hamlyn, 1964), Plates 62, 148, 156, 171, 186, 192, 197, 231, 244, 256, 362, 369.

Also Walter Hamma, Meister italienischer Geigenbaukunst (Stuttgart: Schuler,

1964), plates 603, 604, 615, 622, 630, 640, 642, 653, 654, 656, 659.

8 Heron-Allen, p. 133.

15

This content downloaded from 204.235.148.92 on Fri, 11 Mar 2016 21:43:30 UTC
All use subject to JSTOR Terms and Conditions