Some scholars define French Impressionist composers as those born up to 1902 rather than 1900. This allows for the significant talent of the Debussy of the organ, Maurice Duruflé, who by the mid-20th century contributed to the musical world organ works that define Impressionistic virtuosity—and its most notable Gregorian-based Requiem, with an Impressionistic orchestral accompaniment.
Duruflé is Impressionist in his use of form, harmony, rhythm, and registration. Though organists are thought to favor counterpoint and there are contrapuntal movements, free forms predominate. He emulated his teacher Paul Dukas (1865–1935) not only in compositional style, but also like Dukas he left conspicuously little music for posterity. Beyond the beauty of his compositions, this scarceness may contribute to their being prized, for they are both rare and perfect, and, like the last of anything, especially delicious.
The music is consistently challenging and always rewarding. The melodies, harmonies, rhythm, registration—all grow more loved as they become more familiar, in part because long practice hours are necessary to learn these masterworks. The Suite, Opus 5 of 1934, sets performance demands that have not been exceeded—for the insight to interpret the subtle “Sicilienne,” or the technique necessary for the daunting “Toccata.” His scores are the pinnacle of organ writing, yet in all this remarkable complexity there is never a superfluous note. Duruflé regularly revised the pieces in later life. Of the major works, the Suite, Opus 5, had the fewest revisions from its first publication until the final version. The closing “Toccata” was, however, often the subject of his self-critical eye. He regularly disparaged the piece in masterclasses, never recorded it, advocated cuts, and eventually re-wrote the closing cadenza.
His continual reassessment of his works, his reticence to compose once he began teaching harmony at the Conservatoire National (Supérieur) de Musique in 1943, and his quiet, unassuming manner would have left his music on the shelf were it not for his marriage in 1953 to his brilliant, effusive student Jeanne Marie-Madeleine Chevalier (1921–1999). Her vividly remembered performances and recordings are public testament of her devotion to her husband; her care for him in his infirmity was her private testament. From her début to her final recitals, she performed his works with insight and verve.
Ties that bind:
Opus-to-instrument links in the first editions of the major organ works
The premise
Performance practice studies of other French composers such as Franck or Messiaen investigate connections between their music and the specifications and tone of the organs of which they were titulaire. For Duruflé, information gleaned from masterclasses and suggestions made to private pupils of the composer and of his wife have constituted the basis for performance. Though interesting, this advice given decades after the composition of the works was already stale. Many are unaware that the versions of the scores currently in print were changed from the originals because, unlike most composers who are eager to extend the copyright of their works, Duruflé never renewed his even when pieces such as the Scherzo and “Adagio” from Veni Creator were substantially altered.
For performance today, a careful re-examination of the Duruflé first editions and of the instruments at his disposal when the works were written suggests subtle yet important links between two organs and the compass and registration of the pieces. From 1926 to 1934, the years of his most active organ composition, he regularly played four instruments: in his home town Louviers, the parish church Notre-Dame; and in Paris, the cathedral Notre-Dame, and parishes Sainte-Clotilde and Saint-Étienne du Mont. Of these, he could only play his compositions using his indicated registrations on the organs of the churches he served as titulaire, Louviers and Saint-Étienne. Those where he assisted his Parisian teachers—Tournemire at Sainte-Clotilde then Vierne at the cathedral—were perhaps idealized but are not referenced in the registrations he suggests. These famous instruments are further precluded by their restricted compass.
The background
After a few years of piano and solfège lessons in Louviers, where his father was an architect, in 1912 Maurice went to study in the provincial capital, Rouen. After 1914, he was organist of two Rouen parishes: Saint-Sever and Saint-André, neither with interesting organs. In 1916, his father did design work for the country estate of conservatoire history professor Maurice Emmanuel. After an audition in Louviers, Duruflé began commuting to Paris for lessons with Tournemire to prepare for his entrance into the conservatoire. Emmanuel was a classmate, lifelong friend, and scholar of Claude Debussy. Emmanuel recounts an event he attended in 1887. Théodore Dubois, professor of harmony at the conservatoire, accompanied some verses of the Magnificat at the organ with the then “new” harmonies à la Debussy, using unresolved successions of ninths, elevenths, and thirteenths. Perhaps Emmanuel, a great raconteur, regaled the impressionable conservatoire-bound Duruflé by retelling this incident, or demonstrating the process at the organ, which we certainly hear in Duruflé’s writing.
Once admitted to the conservatoire, he won all the coveted prizes: premier prix in organ with Eugène Gigout in 1922, harmony with Jean Gallon in 1924, and accompaniment with Abel-César Estyle in 1926. That year he wrote his Scherzo to enter the composition program under Charles-Marie Widor. Widor was eventually officially replaced by Dukas, who may have been substituting for him. Duruflé obtained the premier prix of counterpoint, fugue and composition in 1928. In the next two years, he won the improvisation and playing competitions of the French organists’ association, Les Amis de l’Orgue. These dates—1926, 1929, 1930—are critical, for at this time Opus 2 and his two larger works Opus 4 and Opus 5 were in progress on that desk at which he notoriously used the eraser more than the pencil.
The Scherzo, Opus 2, a charming yet intimidating miniature, can be played on a much smaller organ than the large-scale works. The registrations were later changed by Duruflé to a more Neoclassic æsthetic, and the “da capo” was revised. The sonata-rondo form of this piece is exceptional for a scherzo, particularly because of the abrupt changes of tempo. The most remarkable textual variation from the modern version occurs in the da capo, or final reprise of the main scherzo theme. In the 1929 edition, Duruflé quotes the ascending chords from the third theme on the Grand Orgue between each phrase of the scherzo theme. These vignettes were removed in 1947.
The original registrations are more subtly linked to the themes, and sections of the piece are less abruptly demarcated. The effects such as multiple unison stops with Voix humaine, trémolo and sub-couplers are certainly luxurious. Throughout, the ample, embracing original registrations and less pronounced solos reinforce the Impressionistic atmosphere. (See Example 1.)
In addition to registrations, the Scherzo is linked to Louviers by its dedicatee: “A mon cher Maître Charles Tournemire, Hommage reconnaissant.” Tournemire performed the concert for the rededication of the Louviers organ in 1926, at which Duruflé also participated. Perhaps the Scherzo was then played, if only privately. Opus 2 specifies the quiet Récit flute as “Cor de nuit.” Louviers and organs from his Rouen period such as his teacher’s house organ were the ones he knew in 1926 with a Récit flute thus termed. A stop of this name is curiously absent from Duruflé’s monumental specification for Saint-Étienne du Mont of 1956.
Another piece that requests the Cor de nuit is the “Sicilienne” of the Suite, Opus 5, linking it to the Scherzo. In addition to stipulating “Cor de nuit,” the Scherzo and “Sicilienne” require the same manual and pedal compass. Although by 1926 Duruflé was quite familiar with the Sainte-Clotilde organ, its pedal compass precludes that organ. The “Sicilienne” explores a modal, proto-folk melody in the characteristic Sicilian rhythm. The solo appears first in the soprano using Hautbois and Cor de nuit, then in the tenor using Cor de nuit, Clarinette, and Nasard. The second statement is accompanied on the Positif with Bourdon and Dulciane 8'. The “Sicilienne” is further linked to the 1926 specification of Louviers by the request for two 8' strings and a 4' Dulciana on the Grand Orgue, which of all the organs known to him, only Louviers had. In the “Sicilienne,” the Récit “Oboe” is called “Hautbois,” while in the “Prélude” of the Suite, “Basson.” Some may argue that the names are synonyms, but are perhaps unaware that at Sainte-Clotilde, Notre-Dame de Paris, and Saint-Étienne du Mont, the Oboe is called “Basson” or “Basson-Hautbois,” while at Notre-Dame de Louviers, on the house organ of Jules Haelling, and at Saint-Sever, Rouen, the Récit stopped flute is “Cor de nuit” and the Oboe, “Hautbois.” These facts give evidence that the “Sicilienne” may be earlier than the other movements from the Suite and confirm that both works were conceived with Louviers as reference.
Two examples of nonfunctional harmony from the Scherzo and the “Sicilienne” can provide a synopsis of the many stylistic similarities between these two pieces. (See examples 2 and 3: Scherzo measures 181–190, and “Sicilienne” measures 57–61).
Two works linked to Saint-Étienne, as rebuilt in 1928
In 1930 when Prélude, Adagio et Choral varié sur le thème du “Veni Creator,” Opus 4, won the composition prize of Les Amis de l’Orgue, Duruflé was but recently named to Saint-Étienne du Mont, where he had been substituting previously. This was not an instrument such as the monumental, hundred-stop Cavaillé-Coll at Saint-Sulpice played by his fellow Norman Marcel Dupré. Cavaillé-Coll’s Saint-Étienne rebuild of 1873 was succeeded by another in 1883. Renovations were continued by Théodore Puget in 1902 and in 1911 when the Récit was completed with a bass octave. Another rebuild was undertaken beginning in 1928 by Paul-Marie Koenig. This work continued for a time during Duruflé’s tenure, but was abandoned in April 1932. Though mechanically unreliable and unsatisfactory in other ways, Koenig provided 56-note manuals and a 32-note pedal, standard couplers, and a new manual order with the Grand Orgue on the bottom. The only known recital given in these years was by the blind organist Gaston Litaize in March 1931. After 1931 the organ went from bad to worse and ceased functioning sometime before it was dismantled for a rebuild by Debierre in April 1939. From then until 1956, Maurice Duruflé played the Puget choir organ.
The Prélude, Adagio et Choral varié sur le thème du “Veni Creator” is the first work using Duruflé’s familiarity with the organ of Saint-Étienne as reference. A tenor register solo of Récit Clarinette 8' with Nasard is requested. Though the organ did not yet have one in 1928, the replacement of the Cor with a Clarinet was intended as shown in the composer’s specification for Beuchet in 1938, which rebuild would have been with electric action and super-couplers.
On most French organs of the time, pieces like the “Final” of Opus 4 and “Prélude” and “Toccata” of Opus 5 that conclude at the top of the keyboard and request the use of super-couplers in those passages would actually have had no pipes in that range because there were no chest extensions. Therefore, when Duruflé was making requests for super-couplers he had never actually heard them. Perhaps his real-life experience in 1943 of super-couplers with chest extensions at the Palais de Chaillot for the premier of Prélude et Fugue sur le nom ALAIN, Opus 7, convinced him to omit requests for them from the blazing finale of that fugue. One wonders why the sub- and super-coupler indications in Opus 4 and Opus 5 were not among his revisions.
Among many connections linking Opus 4 and Opus 5 to Saint-Etienne are registrations that request Positif at 8' “Principal,” “Bourdon,” and “Salicional” and 4' “Prestant” rather than the generic French names. Saint-Étienne is the only organ he knew offering this precise combination.
Before the introduction of combination action, ventils were used to produce a crescendo. The reeds and mixtures stop knobs were drawn, but the ventil kept them from speaking until a foot lever (appel) was activated. Saint-Étienne was not equipped with a ventil for the Positif because of mechanical limitations. The Opus 4 “Final” begins with Récit mixtures and reeds, Positif foundations, Grand Orgue and Pédale foundation stops with “Anches et Mixtures préparés.” The crescendo calls first for the Positif mixture then separately the reed, followed at Largemente by Grand Orgue and Pédale reeds and mixtures. The drawing of Positif single stops is facilitated by rests in the manual parts, while rests in the pedal allow the use of the appel for the Grand Orgue and Pédale. Because both Notre-Dame de Paris and Saint-Clotilde had a Positif appel, Duruflé’s suggestions in contradiction of standard practice seem clearly intended for Saint-Étienne. Comparison of compass added to the analysis of composite registrations and specifications reinforces the pairings of Scherzo and “Sicilienne” to Louviers and Veni Creator and Suite to Saint-Étienne.
Opus 4 (Veni Creator) was substantially altered in August 1956, and the revised version was issued by the publisher from 1957 onward. There are extensive revisions of the climax, which though treating the same theme, has a much more rhythmically complicated and technically difficult accompaniment. Passages bear a very striking similarity to accompanimental figures in the orchestrated version of the Scherzo. As in the rewriting of the Scherzo da capo, the rewriting of the Adagio climactic section includes removal of interruptions to the crescendo, showing Neoclassic motivation. Tempo and registration revisions seem calculated to make the effect of the piece more homogenized.
A second enclosed division (Positif or Grand orgue) is another curiosity. There was an enclosed Grand orgue on the Haelling studio organ in Rouen where he had lessons in his youth, but he never had one on any organ of which he was titular nor any he designed. Yet, he suggests an enclosed Grand Orgue by a crescendo in the “Prélude” of Opus 4 and a decrescendo in the “Prélude” of Prélude et Fugue sur le nom d’ALAIN, Opus 7.
Although theoretical aspects of the compositional structure of the organ works may exceed the space limitations of this article, similar to the cadence preference of Duruflé in his choral works, third modulations are important. This is clearly demonstrated through the key relationships of the movements of the Suite, which progress by major thirds. The “Prélude” is in E-flat minor, the “Sicilienne” is in G minor and the “Toccata” is in B minor.
In conclusion
The examination of the original registrations of the pre-war works makes clear that when writing them, Duruflé’s model organ was highly influenced by nineteenth-century instruments with strings, celestes and harmonic flutes. His registrations are sometimes generic, such as “Anches” or “Fonds”; however, when specific—“Principal,” “Dulciana 4,” “Cor de nuit,” “Hautbois”—they have been shown to be references to two instruments: Louviers in 1926 and Saint-Étienne in 1928. The composite of the original registrations of the four major works requires 49 manual stops: four 16' flues, thirteen 8' flues, two 16' reeds and six 8' reeds. In the nineteenth century, Barker machines were used to divide the chest between foundations and the reeds/mixtures, which could be controlled by ventils. Cavaillé-Coll and other builders of the late nineteenth century used Barker lever-assisted playing action. The placement of the Positif manual below the Récit and above the Grand Orgue is consistent in all editions of the major works.
Tempo indications were altered in the printed versions. Tempo markings are generally less contrasting in revisions than in original versions. The composer’s ability to update the Scherzo while changing so few notes is quite adept, but the vignettes in the final statement recall the slower themes and intensify his original whimsical concept. The revisions of Opus 2 and Opus 4, and tempo changes, especially taken together with the less warm registrations, lead to this conclusion: in later life he wished them to sound more reserved and matter-of-fact. These “homogenizations” appear to have Neoclassic motivation.
Thus, in many ways, the first edition version of the organ works sounded quite different to his ears and those of his contemporaries when played on instruments of the period with their original Romantic registrations. Their tone and voicing was smooth. Their power was derived from reeds that were rich in fundamental. Unfortunately, most of the instruments Duruflé knew in 1919–1934 have been altered beyond recognition.
In his style there is nothing especially progressive, as one encounters in Stravinsky or Schoenberg. Duruflé was able to manipulate his Ravelian harmonies, Gregorian-like melodies, and contrapuntal textures to go to the very core of the listener’s life. For his is a music that eschews tantalizing the intellect but, in the interest of art, above all pleases the ear. To the end, Duruflé retained the same principles of creativity, which excluded nothing of human warmth. The organs of the turn of the century can inform the performer. The links are too close and too numerous to be coincidental. These are the ties that inextricably bind the works, both the last and the summit of Impressionist organ music, to the late Romantic organ tone for which they were written.
Notre-Dame, Louviers, John Abbey 1887/Convers 1926
I Grand Orgue (54 notes)
16' Montre
16' Bourdon
8' Montre
8' Flûte harmonique
8' Violoncelle
8' Gambe
4' Prestant
4' Dulciana
2' Doublette
Cornet
Plein jeu III
16' Basson
8' Trompette
4' Clairon
II Positif (54 notes)
8' Flûte
8' Bourdon
8' Salicional
8' Unda maris
4' Prestant
4' Flûte douce
22⁄3' Nasard
2' Doublette
8' Clarinette
8' Trompette
III Récit (54 notes)
8' Flûte
8' Cor de nuit
8' Gambe
8' Voix céleste
4' Flûte 4
22⁄3' Quinte
2' Octavin
13⁄5' Tierce
1' Piccolo
16' Cor anglais (free reed)
8' Trompette
8' Hautbois
8' Voix humaine
Pédale (32 notes)
16' Contrebasse
16' Soubasse
8' Flûte
8' Bourdon
16' Bombarde
8' Trompette
Saint-Étienne du Mont, Paris, Cavaillé-Coll 1883/Puget 1911/Koenig, 1928
I Grand Orgue (56 notes)
16' Montre
16' Bourdon
8' Montre
8' Bourdon
8' Flûte harmonique
8' Gambe
8' Flûte creuse
4' Prestant
2' Doublette
Plein-Jeu VI
Cornet V
16' Bombarde
8' Trompette
4' Clairon
II Positif (56 notes)
8' Salicional
8' Unda Maris
8' Bourdon
8' Principal
4' Prestant
4' Bourdon
22⁄3' Nazard
2' Doublette
Fourniture III
Sesquialtera II
8' Cromorne
8' Trompette
III Récit expressif (56 notes)
16' Quintaton
8' Cor de Chamois
8' Flûte
8' Gambe
8' Voix céleste
4' Flûte
4' Salicet
22⁄3' Nazard
2' Octavin
13⁄5' Tierce
Plein-Jeu III
8' Trompette
8' Cor
8' Basson-Hautbois
8' Voix humaine
4' Clairon
Pédale (32 notes)
32' Soubasse
16' Soubasse
16' Contrebasse
10' Quinte
8' Dolce
8' Flûte
4' Flûte
Carillon III
16' Bombarde
10' Trompette-Quinte
8' Trompette
4' Clairon
Saint-Étienne du Mont, proposed specification of 1938
I Grand Orgue
16' Montre
16' Bourdon
8' Montre
8' Flûte harmonique
8' Bourdon
5' Gros Nasard
4' Prestant
4' Flûte
22⁄3' Quinte
2' Doublette 2
Plein jeu II
Plein jeu IV
Cornet V
16' Bombarde
8' Trompette
4' Clairon
II Positif
8' Principal (timbre flûté)
8' Salicional
8' Bourdon
4' Prestant
4' Flûte à cheminée
22⁄3' Nasard
2' Quarte de Nasard
13⁄5' Tierce
Fourniture III
Cymbale III
8' Trompette
8' Cromorne
4' Clairon
III Récit expressif
16' Quintaton
8' Diapason
8' Flûte ouverte
8' Cor de nuit
8' Gambe
8' Voix céleste
4' Flûte
22⁄3' Nasard
2' Octavin
13⁄5' Tierce
1' Piccolo
Plein jeu IV
16' Bombarde-acoustique
8' Trompette
8' Clarinette
8' Basson-Hautbois
8' Voix humaine
4' Clairon
IV Écho expressif
8' Quintaton
4' Principal italien
2' Doublette
Terciane II (Tierce 13⁄5' et
Larigot 11⁄3')
Cymbale III
8' Hautbois d’Écho
4' Chalumeau
Pédale
32' Bourdon
16' Principal
16' Bourdon (ext)
10' Quinte
8' Principal
8' Flûte
8' Bourdon
4' Principal
4' Flûte
Grand Fourniture V
16' Bombarde
8' Trompette
4' Clairon
Composite of registrations of the Scherzo, Opus 2 (III 54/30) and “Sicilienne,” Opus 5b (III 54/31)
I Grand Orgue exp
8' Montre
8' Flûte
8' Flûte harmonique
8' Gambe
8' Salicional
4' Dulciane
III/I, II/I
III/I 16'
II Positif exp (middle manual)
8' Flûte harmonique
8' Flûte douce
8' Bourdon
8' Dulciane
4' Bourdon
III Récit exp
8' Flûte
8' Cor de nuit
8' Gambe
8' Voix céleste
4' Flûte
22⁄3' Nasard
2' Octavin
8' Hautbois
8' Clarinette
8' Voix humaine
Trémolo
Pédale
32' Soubasse
16' Soubasse
16' Bourdon
8' Flûte
8' Bourdon
III, II, I/Péd
Comparison of composite registrations derived from first editions of Veni Creator, Opus 4, and “Prélude” and “Toccata” from the Suite, Opus 5.
Opus 4 Opus 5
56/30 58/31
I Gd. Orgue I Gd. Orgue
Montre 16 Fonds 16
Bourdon 16
Montre 8 Fonds 8
Bourdon 8
Fl. harm. 8
Prestant 4 Fonds 4
Quinte
Fond 2
Mixtures Mixtures
Anches 16, 8, 4 Anches 16, 8, 4
(Bombarde 16)
III/I, II/I 8, 4 II/I, III/I 8, 4
III/I 16 III/I, II/I 16
II Positif exp II Positif
Bourdon 16
Fonds 8
Principal 8 Principal 8
Salicional 8 Salicional 8
Flûte 8
Bourdon 8 Bourdon 8
Prestant 4 Fonds 4
Fond 2
Mixtures Mixtures
Anches 8, 4 Anches 8, 4
Clarinette 8
III/II III/II
III Réc. exp III Réc. exp
Fonds 16
Fond 8 Fonds 8
Flûte 8
Bourdon 8 Bourdon 8
Gambe 8
Voix céleste
Fond 4 Fonds 4
Flûte 4
Nasard
Fond 2
Octavin
Mixtures Mixtures
Anches 16, 8, 4
Trompette 8 Tpt douce 8
Hautbois 8 Basson 8
Clarinette 8
V. humaine 8
Clairon 4
Trémolo
Pédale Pédale
Fonds 32
Flûte 32
Bourdon 32
Fond 16 Fond 16
Flûte 16
Soubasse 16 Bourdon 16
Fond 8 Fond 8
Flûte 8
Bourdon 8 Bourdon 8
Flûte 4 Flûte 4
Anches 32–4 Anches 32–4
(Bombarde 32)
I,II,III/Péd I,II,III/Péd
II,III/Péd 4 II,III/Péd 4
