April 14, 2019


The Meeting of Henry Willis and Aristide Cavaillé-Coll

April 19, 2019


The Meeting of Henry Willis and Aristide Cavaillé-Coll

The 1870 Aristide Cavaillé-Coll organ in Warrington, England at Parr Hall.

The 1870 Aristide Cavaillé-Coll organ in Warrington, England at Parr Hall.

One of the great blessings of being an artist is that we are free to be who we are, free to take whatever inspires us from any other culture in the world and incorporate this into new creations of our own. In music, the world’s greatest composers combined earlier styles with their own new harmonic languages to create timeless masterpieces. Think of Bach, Mozart, Beethoven, Elgar, Howells, Shostakovich, Copland, Bernstein. In architecture, the names of Louis Sullivan, Meis Van der Roh, Walter Gropius, and Frank Lloyd Wright come to mind. In organbuilding, I think of Schnitger, Gottfried Silbermann, Cliquot, Cavaillé-Coll, Sauer, Ladegast, Willis, Harrison, E. M. Skinner, and G. Donald Harrison.

True artists have an appreciation and respect for that which came before them, while at the same time they look toward unexplored areas to make new creations. “What if...” is the question they ask before putting brush to canvas, notes to staff, or lines on drafting paper. Celebrated liturgical architect James McCrery said, “Tradition is necessarily modern and innovative. Only when grounded in historic precedent can architecture intelligently propose and project the future.” I believe we can say this about organbuilding as well.

In organbuilding, the builders of old whom we hold in high artistic regard always put the creation of sound at the forefront of their efforts. As organbuilding styles evolved in the nineteenth century, in which higher wind pressures were necessary to achieve their tonal objectives, the heaviness of direct mechanical actions began to inhibit builders’ ability to create the sounds they were aspiring to achieve. To accomplish their tonal goals, concurrent with the need for their organs to have a responsive playing action, builders invented new actions or assisting mechanisms to reconcile the two seemingly competing forces. In some modern organbuilding circles, we have shifted the focus towards creating a responsive mechanical action first, followed by achieving the best tonal results which the action can allow. The good in this is that organs regained what they had lost prior to the Organ Reform Movement — the ability to play music of the earlier masters — if not “authentically” then certainly more musically than the organs of, say, the 1920s could. Many newly-built organs beautifully reflect upon older styles; however, the development of an original tonal style is limited by the nature of the playing actions. In my journey, I have landed right where Aristide Cavaillé-Coll and Henry Willis did in the nineteenth century: to create a heroic pervading tonal palette, controlled by a responsive playing action.

The historic development of the playing mechanisms using Slider Chests resulted in Cavaillé-Coll’s use of the Barker Lever, Willis’s own Pneumatic Servomotor (an updated version of which is often used by C.B. Fisk) and other far more complicated proprietary systems of the late nineteenth century. These new pneumatic assisting mechanisms for the key actions and the then new-fangled mechanical wind-pumping mechanisms allowed Willis and Cavaillé-Coll to increase pipe scales, wind pressures, and the quantity of wind the pipes consumed, which created tonal results with previously-unimaginable weight and instruments of larger size. All of these organs utilized Slider and Pallet Windchests because these were the kinds of windchests already in use, and therefore the pallets were opened by trackers, which were connected to the pneumatic assisting machine.

The Electro-Pneumatic Pitman Chest was E.M. Skinner’s way of developing a new action to accommodate his distinct tonal style, and to allow consoles to be remotely located from the instruments themselves. Because of the higher wind pressures which his new orchestral tonal style required, and the increasing desirability by many builders to have individual sets of pipes play in multiple locations (unifying, borrowing, extending, etc.), slider chests were no longer practical. Many American organ manufacturers patented their own Electro-Pneumatic actions; the designs (and hence the construction) of these chests became more and more complicated, cumbersome, and ultimately unreliable in the long term.

At my company, we follow the model of Cavaillé-Coll and Willis, and use Slider and Pallet Windchests for three important reasons: inherent tonal blend, tuning stability (both because of the common-note key channels), and mechanical reliability in that there are no exposed moving leather parts. However, we open the pallets by strong electro-magnets which can operate on pressures up to 6” of wind; when higher pressures than that are needed, we install small “helper” magnets behind the pallets to relieve the back-pressure, allowing the large magnet to open the pallet instantly. Our unit stops utilize this very same action, even though there are no sliders in them. In this way every valve opens precisely at the same time, the action is responsive and the repetition is superb.

The solo and liturgical requirements of a nineteenth century (or later) English/Anglican organ are quite different from those in a Parisian Roman Catholic church. The English have a highly-developed tradition of accompanied choral literature, and the French have a rich tradition of symphonic solo organ repertoire. And many American Churches want to do it all. Past attempts at meaningful eclecticism saw organs featuring “German Great Divisions,” French Recit Divisions” and “English Choir Divisions.” The unfortunate result is that these organs don’t hold together as a single creation — they often are not “at unity with themselves.”

We have taken the approach to utilize various national and historical techniques within divisions of the organ. For example, in our larger organs the Great Division has the four elements of the French “fonds d’orgue” but it also tips the hat to the English tradition utilizing 1st and 2nd Open Diapasons. The Swell Division would have a French style Trompette at 8-foot, but the 16-foot Bassoon would be rounder and more orchestral in color. The Oboe could be a plaintive English style “lift-lid” example, which colors the flues nicely for accompanying, or a German style pierced cannister example which can be dark or bright, depending upon how exposed the tone holes are. In recent work we have also included some German romantic colors which intertwine nicely with the palate of sound, completing a holistic approach to our tonal style. In our typical American acoustical environments, this also means modifying the historical prototypes in order that the organ sound in balance with itself and with the building in which it’s installed. Add to that my own personal desire that every one of our organs must be capable of playing anything ever written in a convincing musical way. This prevents our organs from becoming instruments which aspire to everything and then do nothing.

The musical intersection of Willis and Cavaillé-Coll is rather natural. Because of the English Channel, Willis and Cavaillé- Coll were never forced to fiercely compete for work (though several Cavaillé-Coll organs did end up in England). I interviewed Henry Willis 4 in 1994, who recalled that they admired each other’s instruments, and sent pipe-making and voicing samples (specifically harmonic flutes and reeds) across the Channel to each other’s shops. This perhaps explains why the voicing techniques of Willis and Cavaillé-Coll flue stops are remarkably similar. The only significant difference in the pipe-making was that the large Willis zinc pipes utilized a “rubbed-dubbed” lower lip and some “belly” in the pipes (a slightly larger diameter in the pipe’s nodal center than at either end). The rubbed dub was developed to prevent the lips from bowing back into round and spoiling the voicing. Because past generations were mindful not to waste materials, the practice of adding “belly” gave greater structural strength to cylinders made of thinner metal. The net effect on the tone is negligible: a rubbed-dub makes the pipe play a little quicker; belly increases the pipe’s internal area, so the scales need to be adjusted. My colleagues and I have found that by soldering thick lips on the large pipes and scoring the mouth interiors of our smaller pipes, the lips do not move once set by the voicer. Additionally, the tin-rich metal we use to construct our pipes is completely cured and evenly thick, so no “belly” is needed.

Willis and Cavaillé-Coll also had a similar numerical code system for their flue scales: a “No. 1” Diapason scale is enormous (First Open Diapason at Liverpool Cathedral), while a No. 6 Diapason scale borders on string tone. These scales are all translatable into measurements which are not all that different from the scalings my colleagues and I use, as expressed in millimeters. In America, we developed numbered scales, but they are different — an American Scale 40 is a rather large Diapason, and as the numbers increase the pipe diameters decrease. The ratio upon which the pipes’ diameters halve determines the relative strength and color of tone as the musical scale ascends. Today, we cannot simply copy pipe scalings and halving-ratios, but must use variable scalings according to the demands of the project.

There are of course significant stylistic differences between Willis and Cavaillé-Coll which add to each builder’s own personality. The French love splash, éclat, panache, and bright boldness in their full organ sound. This is accomplished to a great extent by their reeds which use parallel shallots and long tongues in the bass octaves. The English love fullness and richness, but more in the context of a mature cheddar and sherry, rather than piquant blue cheese and pinot noir. Their roundness of tone results in part from darker reeds which use tapered shallots and weighted tongues in the bass.

In order for the flue ranks in a Cavaillé-Coll organ to maintain their potency with the reeds, their scalings are much larger in the treble than English organs. In this way the flues maintain their strength with the wash of brilliant reed tone in the reverberant acoustics which naturally attenuate treble frequencies. In Willis organs, the full sound does not depend upon large scaled flue trebles. However, the full English sound does often include harmonics mixtures in which third and flat-seventh ranks are introduced, which mirror and amplify the harmonic structure of the darker reeds, to give one the perception that the flue pitches are not lost. We strike a balance between the two.

Willis mutations are much smaller scaled than Cavaillé-Coll’s, and therefore have a distinctive bright quality about them, even when voiced softly. (Henry Willis 4 likened this to using a pinch of Coriander, rather than a tablespoon.) The upper lips of Willis harmonic flutes are made in the round of the pipe body’s cylinder and highly arched, rather than flat.

In my experience, a Willis style 4’ harmonic flute lives nicely alongside a Cavaillé-Coll style 2’ octavin or 8’ traverse flute. The principal choruses of both builders are remarkably similar in their effects in large spaces. Apart from some low quint pitches in some Cavaillé-Coll mixtures and some third-sounding ranks in Willis mixtures, the overall “clang” of the two full flue choruses are really quite compatible with one another.

In the recent French Flue Voicing Seminar presented by the American Organ Institute at the University of Oklahoma, students were taught how to take flue-pipe voicing samples from the Cliquot era and turn them into Cavaillé-Coll examples. Many of the students remarked as to how similar the Cavaillé-Coll voicing style is to many of modern American builders’ practices.

Both Willis and Cavaillé-Coll flue stops consume similar quantities of wind, and it’s the quantity, not just the pressure of the wind which assists in giving a flue pipe a pervading all-encompassing tone quality. This gives a full accounting of the harmonic series even in softer open stops, to carry the fundamental sound into large spaces and to bind to the other stops. If you listen to examples of Willis and Cavaillé-Coll flue voicing blindfolded, you might be challenged to tell which is which. For example, the flue-work sound of the Willis Organ at Truro Cathedral is remarkably similar to that of St. Sulpice in Paris.

It is in the reeds and scaling of the flue trebles where each builder makes its individual statement; through them, concurrent national styles of organbuilding and musical composition evolved into specific schools. In French organs, the blazing Trompettes and big scaled flue trebles give the full organ sound a thrilling personality and hence, their national identity. Since English organs were focused more on choral accompanying, softer orchestral reeds were developed to blend with the flues and for cantabile, obbligato passages, imitating more closely a modern orchestra. But this does not mean Willis organs are without exciting reeds. The Willis full organ “clang” tips its hat to what he learned from Cavaillé-Coll and the French style — trumpets and harmonic trumpets with parallel shallots. Willis added weights to his tongues to darken and solidify the bass, whereas Cavaillé-Coll did not.

True French reeds are not simply bright, but rich in fundamental as well. The old adage is that one should perceive a “Bourdon” sounding in every Trompette note. This is created in part by large scales of the resonators — which both Willis and Cavaillé-Coll understood — rather than the small-scale resonators we see in many American organs from the mid-twentieth century. This is a mistaken notion of how to adapt French reeds to American acoustics — the tone only becomes thin. Henry Willis III wrote to G. Donald Harrison following a trip to the United States in the 1950s and warned him not to decrease the scales of the Trumpet style reeds as other American builders were beginning to do. Regrettably, that admonition fell on deaf ears.

Willis and Cavaillé-Coll utilized these techniques of design, pipe-making, scaling and voicing in their instruments to a very similar degree. Both of them developed playing actions to accommodate development of the primary artistic aspect of an organ — its sound. In following these precepts, my colleagues and I build instruments which are able to register several different seamless crescendos leading towards full organ. These principles are why a Willis Organ can accompany the Duruflé Requiem, a late Cavaille-Coll Organ can accompany an Victorian Anthem and a Buzard Organ can go from Bach to Duruflé to Howells without breaking a musical sweat.


John-Paul Buzard is the President & Artistic Director of Buzard Pipe Organ Builders, LLC of Champaign, Illinois. He received the Master of Music degree in Organ and Church Music from Northwestern University in 1980, and his Master Organbuilder Certificate from the American Institute of Organbuilders in 1985.

The Buzard Organ Company is housed in a converted turn-of-the-Century hotel in downtown Champaign and in a storefront and warehouse a block away, and employs 20 persons, all of whom are Members of the American Institute of Organbuilders. John-Paul is also a member of the American Guild of Organists, The Organ Historical Society, and the International Society of Organbuilders, and his firm is a member of The Associated Pipe Organ Builders of America.

John-Paul is also a Member of the Worshipful Company of Musicians of the City of London, and a “Free Man of the City of London,” having been sponsored by organbuilder Henry Willis, 4 as the only non-English resident ever allowed into this ancient musicians’ craft guild. He is also the first and only recipient of the Henry Willis Organ Pipe Voicing Medal, awarded by the Worshipful Company of Musicians following demonstration of his abilities according to an established examination protocol.

The views and opinions expressed in this article are those of the author, and do not necessarily reflect the position of Vox Humana.