The term used to denote collectively the various mechanical contrivances which have been devised to supplement the fingers in controlling the tone holes of wind instruments. The function of a key is to enable finger pressure applied at a convenient point to open or close a hole of any required size in any required position. Without the prior existence of established principles of keywork, some modern instruments designed following rational acoustic lines, for example the saxophone or heckelphone, could not have been realized.
2. Early form and construction.
3. Modern form and construction.
4. 19th- and 20th-century systems.
5. Experimental and unusual systems.
PHILIP BATE/JERRY VOORHEES
From very early days the need has been felt to provide wind instruments with a more extensive and musically useful scale than the mere harmonic series proper to a tube of fixed length. The process of varying the ‘effective length’ of a tube by means of side holes opened and closed by the fingers has been used empirically since the neolithic period, as shown for example by bone pipes preserved in burial places, but the systematic disposition of such holes seems to belong to a much later cultural stage. The oldest surviving examples of organized side hole arrays known at present are of Sumerian origin, and date from about 2800 bce. Thereafter deliberately positioned side holes are found in instruments of successive cultures up to the eclipse of the Roman Empire and the coming of the Dark Ages, but there is no sure evidence among them of any device in the nature of a key to supplement the fingers, unless the mechanisms used on some Greek auloi to close off unwanted holes when playing in specific modes are accepted as such (see Aulos, §I, 5(v)).
From the 12th century ce onwards there is some evidence of both reed pipes and flutes with six holes giving a diatonic scale of seven degrees, the octave being sounded by overblowing the lowest note. This organization also provided certain more or less satisfactory intermediate tones by the process of ‘fork fingering’. In addition, the overall compass was sometimes extended by lengthening the tube and boring a further hole which could be stopped by the lowermost little finger. By the early 16th century such pipes were being made in various sizes paralleling the different ranges of the human voice, and it was at this stage that practical difficulties began. The acoustic laws relating the size and position of side holes made possible the placing of the six primary holes in two fairly close-set groups of three, but this device could be carried only so far. By simple proportion the longer the tube, the wider the spacing between holes necessary for reasonable intonation, soon exceeding the stretch of the normal hand. On some instruments the player can adjust the pitch of certain notes by blowing; when the limit of this ability is reached mechanical assistance becomes essential.
The first primitive keys were most likely designed to assist the little finger; the oldest surviving authoritative illustrations, by Virdung (1511) and Hans Burgkmair (Maximilian's Triumphal Procession, c1516–18), show no other arrangement. Accounts of the Duke of Burgundy for 1423 and 1439, recording the purchase of ‘bombardes à clef’ and ‘teneurs à clef’ for the court, show that keys were in fact used nearly a century earlier. Teseo Ambrosio's account of the Phagotum offers some proof that a more advanced application of simple keys existed by the 1530s, and Praetorius in Theatrum instrumentorum (1620) shows keys for the thumbs on both large shawms and bassoon types. From this period on there is evidence of an increasing use of simple mechanism, although at the end of the 18th century it remained somewhat crude and often inefficient. As late as 1815 Gustave Vogt, professor of oboe at the Paris Conservatoire, questioned the efficiency of keys and advocated using as few as possible. At that time instrument makers in Germany and Austria were already quite generous in providing them.
At least three centuries seem to have elapsed between the invention of keys for wind instruments and their application to improve the layout of the primary holes. Borjon de Scellery, in a celebrated plate published with his Traité de la musette (1672), depicted a large oboe (possibly a quinte de hautbois; see Oboe, §II, 4, fig.3) which appears to have something of the sort, and a few similar instruments are preserved in continental museums as ‘basses de musette’, but these are thought to have been exceptional and their history is obscure. A bass transverse flute by Jan Beuker in the Paris Conservatoire collection has two jointed open keys which much improve the primary layout; the instrument dates from the first half of the 18th century. Similar instruments employing simple second-order levers were illustrated in Diderot's Encyclopédie in 1769 (‘Lutherie’) and were being made by Delusse of Paris in about 1760. In 1810 the London maker Malcolm MacGregor patented an instrument of this type and the principle was revived again by Abel Siccama of London in the 1840s. These apparently unrelated recurrences of a single idea prove the difficulty of tracing clear and unbroken lines of development (if indeed such exist) with keywork as with many other features of musical instruments. There is unfortunately no authoritative instrumental historian between Martin Agricola (published 1528 and later) and Praetorius (1619), a period rich in important developments. It is well to keep in mind that through natural conservatism or prejudice new devices were often slow to replace older, less efficient ones, and to avoid reading into slender evidence more than is warranted.
The use of keys to create a chromatic scale probably began about the last quarter of the 18th century, firstly with the transverse flute. Before that time resistance fingerings (fork fingerings and half-hole fingerings) had furnished all the primary semitones except the lowest and with reed instruments, where this technique was quite satisfactory, it continued to do so for many more years; even today this process is of much service. Modern research has explained why the acoustic properties of the one-key flute make it more difficult to ‘pull’ into tune when resistance fingerings are used than was, for example, the 18th-century oboe, and no doubt this deficiency stimulated experiment with chromatic keys.
The earliest known closed keys were simply cut out of a single piece of sheet brass, with the touch, shank and pad cover all incorporated into the unit (fig.1a). The pad cover was faced with leather either sewn or cemented on. Early open keys assumed two configurations, the more common being similar to the closed key except that it required two interlinked shanks, each having its own transverse pivot (fig.1b). The other type was a second-order lever, with a pivot on one end of the shank, a padded plate on the other and a touchpiece in the middle. In both cases the pivot was a wire passing through upward-bent and perforated lugs. This arrangement was somewhat delicate and was provided with protective covers of wood or metal in almost all the earliest examples (see Shawm, fig.1. In the later 17th century the key shank lay in a slot cut in a ring, later reduced to a block left standing above the surface of the body tube. A small hole bored tangentially through both ring and shank carried a close-fitting pivot pin. This arrangement is in fact efficient, if inelegant, and remained in use until the early 1900s (fig.1c).
In the early 18th century metal saddles of channel section, screwed or pinned to the joint, began to augment or replace the wooden mounting blocks (fig.1d). Their principal use was on bassoons as it was difficult to leave a wooden ring or block standing when carving a large, irregularly shaped joint, but metal saddles also appeared on oboes and clarinets, often in combination with the older wooden blocks or rings. The existence of both types of mountings on the same instrument does not necessarily indicate that the saddles were later additions.
The tightness of the seal between pad and joint depended on the resilience of the facing material and the shape of the pad seat. Fashioning the earliest pad seats involved merely flattening the tube surface around the hole, but by the end of the 17th century in France pad seats were more developed. The form of the padded plate and of the touchpiece sometimes helps to date woodwind instruments. The earliest known padded plates were round or racquet-shaped. In France at the end of the 17th century they were either round or rectangular, and by the end of the 18th century octagonal plates were fitted by many oboe makers who used characteristic decoration (see Young). The ‘fishtail’ shape, seen on some touchpieces (fig.1b), may indicate an origin before the firm adoption of the ‘left hand above right’ playing position, although on oboes it survived as an ornament perhaps as late as 1830. Early return-springs consisted of a leaf of hard brass fixed to the surface of the instrument body and pressing upwards against the key body. Later a spring of brass, then steel, was screwed or riveted to the key itself and, to reduce friction and wear, bore on a slip of metal set into the surface of the body tube. When nearly all woodwind keys were simple, they were named according to the note sounded when the finger was applied. With more modern, integrated fingering systems this system of nomenclature has become inadequate.
A device fundamental to all modern keywork made its first appearance in about 1800 on instruments of high quality. A pair of turned metal pillars is attached to a fitted metal footplate or screwed directly into the tube wall. The pivot associated with these pillars may be either a tube threaded on to a steel axle supported between the pillars (fig.1e), or a solid rod supported by conical pivot screws, one in each of the paired pillars (fig.1f). About 1830 a most useful variant of key design appeared, in which the padded plate and touchpiece were attached to the opposite ends of a long rod or tube pivoted between pillars as described above. When both the touchpiece and padded plate extend from the rod in the same direction, the key is an open key (fig.1g). When they extend in opposite directions, it is closed (fig.1h). Keys of this sort were a great improvement over long levers where wide stretches had to be bridged; since a key of the new type was anchored at each end by a pillar, there was less chance that the position of the touch or pad would be disturbed by a flexing of the shank.
By the mid-18th century cast or forged bar metal had begun to replace sheet metal for key shanks, but brass remained the usual metal for keywork through the early 19th century, with silver at times for superior instruments. White bronzes – generically known as German silver or nickel silver – were introduced about 1830 and became the standard material for modern keywork. In the late 19th century the hand production of standard-pattern keys in quantity flourished as a home industry in France, contributing much to the ability of French makers to sell instruments of decent quality at highly competitive prices.
Another advance was the invention of the needle spring (fig.1i), most probably by Auguste Buffet jeune in Paris about 1838: a tempered wire of gold or steel is anchored in one of the pillars, while its free end bears on a tiny hook soldered to the key tube. The bias of such springs is very easily adjusted and they are widely used today on instruments of all sizes.
Pad airtightness was considerably improved around the middle of the 19th century by the invention of stuffed pads, or ‘elastic plugs’ as some makers called them. These were ‘purses’ of fine kid, filled with a ball of lamb's wool and drawn together with a thread. Like the natural fingertip, they were flexible and adapted themselves to seal the holes. As the pads could not be easily attached to flat cover-plates with shellac, the cement in common use, the key-heads were themselves modified to a hemispherical cup or ‘saltspoon’ form (fig.1j). But stuffed pads tended to bulge in use and ‘shade’ the tone hole, flattening the pitch and damping the tone. The modern pad thus comprises a cardboard disc, a layer of felt and a covering of fine kid, animal membrane or waterproof plastic film, supported in a much shallower, flatter cup. Some modern pads are made of resilient, homogeneous plastic, and many professional oboes are padded mainly with cork. Pad seats have developed into carefully cut recessed cones on non-metal instruments (fig.1k) and raised collars in the walls of metal ones (fig.1l).
During
the middle third of the 19th century improved acoustical knowledge,
manufacturing innovations and escalating performance demands inspired
revolutionary developments in keywork, especially interdependent systems in
which keys acting on different tone-holes were linked. New devices addressed
the same tasks to which earlier, individual keys had been applied: bridging the
distance between the sidehole and its operating finger, extending the range and
replacing or improving fork fingerings. The brille (Ger.: ‘spectacles’), a
device based on a ring surrounding a side hole, helped to equalize the tone of
fork-fingered notes. Although Frederick Nolan apparently used such rings as
early as 1808, the true potential of the device was not realized until Theobald
Boehm combined a pair of rings with a satellite pad cup on his flute of 1831 (fig.1m). By allowing three fingers to control
four open holes the brille provided greatly improved venting for fork-fingered
notes, and gave inventors the means to devise logically inspired fingering
systems with far less regard for the physical limitations of the hand. The
clutch uses levers with interlinked ends to modify key action. Early open keys
used a clutch to connect the two key-shanks, one of which carried the touch and
the other the pad cup. In this application the height adjustment of the clutch
was not critical. About 1840 Buffet jeune assigned new responsibilities
to the clutch by creating a mechanism in which both lever arms carry a pad cup.
In this case the height of the clutch must be carefully adjusted by shims or an
adjusting screw (figs.1n and 1o) to ensure that both pad cups close
their associated holes simultaneously and completely. Such a clutch may be
termed critical. By proper use of critical clutches many complex actions can be
devised: for example, the function of a brille can be duplicated in a mechanism
which uses closed pad cups instead of rings (fig.1p). The use of opposing springs was
another mid-century innovation. Typically, this mechanism involves a weak
opening spring fitted to a satellite pad cup which is normally held closed by
one or more levers equipped with stronger closing springs. Useful results may
be obtained by applying both opening and closing pressures simultaneously. One
of the earliest applications of this device was the G
mechanism for flutes (fig.1q) devised by Vincent Dorus and Louis Lot
in 1837–8 (see Boehm, Theobald). Frédéric Triébert's système 5
oboe of about 1855 provides a possibly more useful example (fig.1r). Two subordinate mechanisms complete
the list of devices used in modern fingering systems: the rocker, which
resembles a seesaw (fig.1s), and the bridge bar, which carries
motion around intervening mechanisms mounted on the same axle.
The range of a woodwind instrument is extended by lengthening the tube at the lower end to add lower notes or by facilitating the production of the higher harmonics in order to add or stabilize higher notes. In both cases keywork is essential. The note produced when all six primary holes are covered is called the six finger note and marks the lower end of the domain of the primary fingers. Notes below that comprise the extension: these are usually controlled by the little fingers of both hands, using keywork. On bassoons and some low-pitched clarinets the added tube is so long that the thumbs must also be employed. The challenge faced by builders devising keywork for the extension was to use two comparatively weak fingers to control several notes (at least three, usually four or more) in a sure, smooth manner. Keywork for the extension has generally involved simple keys and clutches, and sometimes opposing springs. Smoothness was achieved by providing duplicate touches so that the fingers could alternate, by providing rollers between touches so that they could slide gracefully from key to key, by using opposing springs so that pressing certain keys simultaneously gave the desired result, or by any combination of these. There is more variety among extension mechanisms than among keywork controlled by the principal fingers, and this variety seems to be based on the traditions of the individual woodwind types rather than on logical considerations. Although the demands on the extensions of modern woodwind instruments are similar, every woodwind type retains its characteristic extension design.
On all woodwind instruments except flutes the production of higher harmonics is usually aided by small vents high on the instrument's body, controlled by keys operated by the left thumb. These octave or register keys are often simple closed keys. One is usually sufficient for clarinets, while bassoons may require three or more. Many modern oboes, most bass clarinets and nearly all saxophones have an automatic register mechanism which, when activated by pressing the left thumb touch, chooses between two register vents on the basis of linkages with elements of the lower keywork. There are many variations of this ingenious mechanism but most require clutches, rockers and opposing springs (see fig.1s).
19th-century
fingering systems that have been widely adopted have represented one of two
general approaches. One type, exemplified by the Albert-Oehler clarinet and the
conservatoire system oboe, evolved pragmatically as mechanical devices were
added to enhance the capabilities of the six traditional finger-holes
(evolutionary). On the other type, over which Theobald Boehm had a great deal
of influence, a geometrically correct series of holes for the semitones was
calculated and a suitable mechanism devised to control them (Boehm system); for
example the Boehm flute and the saxophone. With the exception of bassoons all modern
woodwind instruments of both types have a brille controlled by the primary
fingers of the right hand. The type of brille used helps to identify which of
the two approaches a particular instrument represents. On evolutionary
instruments the rings of the brille surround the lower two holes and the
satellite pad cup lies between the two upper holds. When the three principal
fingers of the right hand are lifted consecutively from the bottom, the
intervals produced are tone, tone, semitone. On oboes and similar instruments
the notes produced are D, E, F and G, and the satellite vent produces a fork F with
a tone approaching that of the other pitches. Such instruments invariably have
a prominent closed key for F which lies between the lower two holes (fig.1t). On Boehm system instruments, raising
the right-hand fingers consecutively gives tone, semitone, tone: that is, D, E,
F, G on flutes and similar instruments. The rings of the brille surround all
three finger-holes, and the satellite pad cap is located above the highest
finger-hole where (when closed by the lowest ring) it improves the fork F (fig.1u). Most modern flutes, oboes and
saxophones now have padded plates instead of rings (see fig.1p).
Beside the mechanisms which have gained general acceptance there is an interesting group of experimental and rejected mechanisms, many of which show remarkable ingenuity. In general, these were attempts to solve some perceived problem which the standard mechanisms do not, in the mind of the inventor, sufficiently address.
Boehm
proposed that ideally every semitone on a woodwind instrument should have its
own hole, that the hole should be correctly placed geometrically, as large as
practically possible and associated with an open-standing key. The realization
of this concept is termed full venting. Modern flutes and saxophones approach
this ideal, but practical considerations prevent its full implemation. Attempts
to bring the flute's mechanism closer to the ideal have involved the addition
of open G and D keys, and a separate vent
for C. Boehm himself
considered improvements in all these areas and was a strong advocate of the
open G. This
represents one of the few instances in which Boehm's idea was not accepted by
modern flautists. He also considered an open D, but felt that having the player hold the D key open was better
because it stabilized the playing position. Inventions by others provided open
G and D keys, but at the cost of illogical
and unwieldy control of the instrument's lowest notes. Some inventors also
attempted to improve the venting of e''' and f''' by devising various complex
mechanisms. Those working along these lines included J.C.G. Gordon, R.S.
Rockstro, Cornelius Ward, T. van Everen, François Bourne, Djalma Julliot and,
more recently, Alexander Murray.
Several attempts were made to apply Boehm's concepts of keywork and full venting to instruments other than the flute. Buffet jeune transferred much of Boehm's keywork to the clarinet but wisely refrained from adopting the concept of full venting; his instrument (patented 1844) has become the standard. Several builders added elements of Boehm's keywork to oboes and bassoons, with or without using full venting. Perhaps the best-known attempts were made by the Triébert firm in Paris, for which Boehm provided (at least in the case of the bassoon) the geometric schemata for tone hole size and placement. Elegant and complex, Thiébert-Boehm system oboes and bassoons still exist in museums but their tone quality is deemed uncharacteristic, probably due to the brightening effect of the large, open-standing tone holes. Along similar lines, various saxophone-fingered oboes were tried. Builders associated with these experiments include Adolphe Sax, H.J. Haseneier, A.J. Lavigne, John Sharpe, Ward, C. Kruspe and the firms Gautrot, Triébert, Boosey & Hawkes, Buffet and Heckel.
A number of flute makers desired to obtain the acoustical benefits of full (or at least increased) venting while retaining, in essence, the fingerings of the classic instrument. Builders working along these lines included Siccama, Richard Carte, R.S. Pratten, Giulio Briccialdi and John Clinton. Some later designs from Rudall, Carte contrived to put both Boehm and classic finger patterns on the same instrument, letting the player decide which to use. An interesting inversion of the movement towards full venting is the existence of keywork designed to decrease the venting of certain notes in order to preserve, as much as possible, the classical venting. An example may be found in the right-hand mechanism of Viennese oboes by Zuleger.
Generations
of amateur clarinettists have been annoyed by the need to move the left index
finger from its home position over the first finger-hole in order to play the
closed keys for g', a' and b
', and several builders have invented
keywork that controls these throat vents without displacing this finger. These
keywork systems connect the vents to rings surrounding the primary side holes
of either or both hands; they are usually complicated and, although they can be
made to work, none has ever achieved much acceptance by professional players.
Those engaged in devising this sort of mechanism include V.-C. Mahillon, the
firm of Rudall, Rose, Carte & Co., Antonio Romero y Andía, G.H. Child and,
in the middle of the 20th century, T.F. McIntyre.
A small but remarkable set of woodwind instruments exists on which a semitone or chromatic series of changes is produced as the principal fingers are lifted consecutively, instead of the usual diatonic progression. The goal of the inventors was to eliminate all (or nearly all) keywork. The difficulty with this concept lies in the unalterable fact that humans are not endowed with enough fingers to control a 12-note chromatic scale without keywork or fork fingerings, to say nothing of supporting the instrument. Nevertheless, C.T. Giorgi produced, at the turn of the 20th century, an 11-hole, keyless, chromatic vertical flute which required the use of every finger and thumb as well as the side of the left index finger for musical purposes. Other inventors augmented or replaced some of the open holes with keys, but the basic chromatic fingering pattern which defines instruments of this type has made them generally unacceptable to professional musicians. Other inventors who have worked on such instruments incude Siccama, Ward, R. Burghley, and H.L. Schaffner.
See Fingering, §III, 1–2.
Waterhouse-LangwillI
YoungHI
T. Boehm: Die Flöte und das Flötenspiel (Munich, 1871/R; Eng. trans., 1922/R)
V.-C. Mahillon: Catalogue descriptif & analytique du Musée instrumental du Conservatoire royal de musique de bruxelles (Ghent and Brussels, 1880–1922, repr. 1978 with addl material; i, 2/1893; ii, 2/1909)
R.S. Rockstro: A Treatise on the Flute (London, 1890, 2/1928/R)
A. Carse: Musical Wind Instruments (London, 1939/R)
F.G. Rendall: The Clarinet (London, 1954, 3/1971 by P. Bate)
P. Bate: The Oboe (London, 1956, 3/1975)
A.C. Baines: Woodwind Instruments and their History (London, 1957, 3/1967)
L.E. Gilliam and W. Lichtenwanger: The Dayton C. Miller Flute Collection: a Checklist of Instruments (Washington DC, 1961)
L.G. Langwill: The Bassoon and Contrabassoon (London, 1965)
P. Bate: The Flute (London, 1969, 2/1979)
P. Bate: The Bate Collection of Historical Wind Instruments (Oxford, 1976)
N. Toff: The Development of the Modern Flute (New York, 1979/R)
J. Voorhees: The Classification of Flute Fingering Systems (Buren, 1980)
G. Joppig: Oboe und Fagott (Berne, 1981; Eng. trans., 1988)
W. Waterhouse: The Proud Bassoon: the Waterhouse Collection of Bassoons and Related Items (Edinburgh, 1983)