Category: Piano Repertoire

Cecile Chaminade: ‘Marine’, op. 38

Like many of Chaminade’s solo piano works, Marine, op. 38, is in ternary form—the A section being repeated literally before leading to a coda. Gratefully written, it exploits the full range of tone-colour offered by the piano, from the sonorous depth of the bass to glistening arabesques in the upper register.

The title, somewhat vaguely translated as ‘Sea-Piece’, effectively portrays a calm, rolling ocean. Here there is little sense of peril such as that in Ravel’s much more overtly dramatic ‘Une barque sur l’océan’, from Miroirs. The pervasive trochaic rhythmic creates a sense of rise and fall, the low-register bass suggests vast depths of water beneath us, while the pedal-notes lend stability. Of the A section, only one bar abandons the instrument’s lowest octave. Above this, the right hand contains faster-moving broken chord figures and rippling arabesques, little waves splashing on the water’s surface. Only in the coda do things take a more ominous turn, but this threat is short-lived, and clear skies and blue water by and large prevail.

Chaminade: Marine, op. 38, bars 1-4

The A section presents the main idea—though it can hardly be called a melody—the right hand elaborating in shorter note-values a rising-and-falling shape in the left hand. The harmony, essentially oscillating I-V7, is underpinned by a tonic pedal until halfway through bar 5, when the move to the dominant, which arrives at bar 7, is initiated. To provide some relief from the I-V harmonic bias, Chaminade makes colouristic use of an (implicit) E𝄬7 chord, sometimes with a diminished fifth, functioning as German/French 6th or Neapolitan 2nd, depending on the prevailing key. First in bar 4, then in bar 8, where it tugs against the dominant pedal in the bass, it occurs again in bar 10, when the bass compromises, and the music briefly drifts towards B𝄬, before sliding chromatically back to the tonic.

It is worth observing the sense of breadth, of expansiveness, that comes from Chaminade’s consistent deployment of over four octaves of the piano’s compass. If we think in terms of G major, this compass is six octaves (most instruments lack a low ,,G). The music avoids the highest available octave until the final flourish in the penultimate bar. Likewise, the music only rarely descends below the lowest available tonic: once, establishing the dominant in bar 7 (and bar 40 in the A-section reprise), then the flat submediant in bar 10; there are a couple of low dominants in the relative major of the B section (that is F, the dominant of B𝄬, the relative of G minor); the (global) dominant in bars 28 and 34; and one flat submediant in bar 32.

The Aeolian-tinged B-section melody, in the parallel minor, has something almost shanty-like about it. While maintaining the trochaic rhythm, repeated block chords replace the rising-and-falling arpeggios, which creates a greater sense of solidity, as does the increased concentration on the piano’s middle three octaves. Having modulated to the relative, B𝄬 major, by bar 21, a brief reminiscence of A-section figuration (bar 23) instigates a hushed return to G minor in bars 24-5. The ‘bleeding’ of material from one section into another also creates a more fluid sense of structure, unifying the whole.

Chaminade: Marine, op. 38, bars 18-21

Note how Chaminade extends by one beat the imperfect cadence in bar 27—this reinforcement of the tonic occurring at the point where the original statement of the material began its modulation to the relative major. This causes the metrical displacement of the answering phrase, and the resulting one-beat extension of the close into B𝄬 major at bar 30. Masterful, too, is the re-use of the material from bars 22-3 in bars 31-2, now centred on V and 𝄬VI6dim5 of G then, V and V+, to usher in the A-section reprise.

This reprise is literal until bar 48. Bar 50 is a repeat of bar 16, though with a varied left hand. Bar 49, while based on bar 16/50, delays the expected perfect cadence—marking an upward 6^-7^ progression. This cadence, however, is interrupted by—again!—the E𝄬 German sixth, in the most sombre turn of the piece, against an ominous dominant pedal, tolling in the deepest register. Bar 55 resolves this into a G+, then to G (the augmented fifth remaining, enharmonically, as an E𝄬), before all the gloom is dissipated in a shimmering trill and upward flourish—one final spray of foam. Finally, the low, rolling chords of the opening return to bring the piece to its conclusion.

For me, the remarkable aspect of this work is just how evocative it is, and also the economy with which Chaminade is able to construct a cohesive and cogent whole from a small number of brief ideas. At its worst, musical tone-painting can be little more than a string of pictorial effects. Such a piece must work on two levels: first, the musical devices employed, while not resorting to mere mimicry, must indeed convey a sense of the image, mood, atmosphere, or whatever, that the composer intends; second, and most importantly, the piece must make as much sense, purely as a piece of musical composition, to a listener who is unaware of the composer’s intended representation as it does to one who is aware of it. This movement succeeds magnificently in both regards.

From the pianistic perspective, the focus on rolled chords, pedal notes, and arabesques, makes this as joyous to play as it is satisfying to listen to. Good, reliable fingering in the right hand figurations, of course, is essential for success. While many of the left-hand chords are tenths, a big hand is not essential—good use of the wrist is. That said, it’s not only acceptable, but perhaps better tone-painting, to take a little more time over the rolled chords. As long as they’re not so slow that each note sounds as if it’s rhythmically distinct, it’s quite within the feeling of ebb and flow.

Dynamics should be on the subdued side: only six forte and four fortissimo markings, as opposed to ten pianos and thirteen pianissimos (and even a ppp); the main A-section idea is marked dolce, and dolcissimo at the reprise; and the two arabesque passages (bars 13 and 41) and the final trill are all marked leggierissimo (calling for delicacy, rather than brilliance). Although the B-section melody is marked marcato on the four occasions it appears, I would suggest that a more sympathetic interpretation of the word here—rather than the usual ‘accented’—is ‘projected’ or, in other words, at a distinctly higher level of tone than its chordal accompaniment. We can perhaps induce from this that there is not one dominant melodic voice, that the A-section material can be voiced more equally. 

Use restraint, too, in the various requested tempo fluctuations: Chaminade is fastidious about whether the rit. is poco or molto, so observe the differences carefully, but remember, too, that the relentless swaying of the rhythm is fundamentally important to the character of this piece.

Chaminade’s pedal markings are also detailed, and there are moments where she specifically requests use of the left pedal, too. Remember, though, that all instruments and acoustics are different and what will be effective in one circumstance may not be so in another. In this regard, we should always try to honour the spirit, rather than the letter, of the score.

Franz Liszt: La cloche sonne

Although I rarely have cause these days to paddle in the balmy waters of the lower grades of the ABRSM syllabus, a little piece on the 2023-24 Grade 4 set works list caught my eye, and so I wanted to share my thoughts on it, for various reasons. First, of course, it’s a nicely-crafted little piece, which is full of potential for atmosphere and feeling for students to enjoy exploring. Second, and not insignificant, is that it is by Franz Liszt, who is one of my favourite composers, but who is also not a chap who’s well-known for making things easy and, as a result, I think this is only the fifth time I have ever encountered his work on a grades syllabus. Sadly, it is an ‘alternative’ piece on the syllabus, meaning it isn’t printed in the exam pieces collection, so without a little more initiative on the part of teacher or student is more likely to be overlooked, so my third reason is promotional.

La cloche sonne nicely exemplifies one of my favourite little sayings: a single manuscript can change the world. It was first published as a piano piece in 1958, from a manuscript originally in the Liszt Museum and now in the collection of the Goethe- und Schiller-Archiv in Weimar, Germany (shelfmark GSA 60/Z 1). Thanks to their wonderful digitisation project (an initiative blissfully well-funded and widespread throughout European research libraries) it is now available online, under a public domain licence, which means I can include it here.

RV 96 La cloche sonne Altfranzösisches Lied Auf der Rückseite Klaviersatz. Liszt, Franz

And here are my diplomatic transcriptions:


(Obviously, there are no bar numbers in the manuscripts; I’ve added them for ease of reference.)

Previously, the work was believed to be based on an unknown French folk song, and is still catalogued as such, for instance, on IMSLP. No-one has ever been able to trace the folk song itself, and this is where the ‘changing the world’ happens, because another manuscript appeared for auction at Sotheby’s in 2019, and was subsequently published by the Liszt Society. There is a fabulously clear image of the manuscript on their website.

https://www.sothebys.com/en/auctions/ecatalogue/2019/rbl-pf1923/lot.86.html

And here is my diplomatic transcription of this version:

The inscription at the bottom reads: ‘L’Angelus – mélodie de Madame la Baronne Amable de Béville, notée par son très humble serviteur’ (‘The Angelus – melody by Madame la Baronne Amable de Béville, notated by her most humble servant’), and signed by the composer. This strongly suggests that it is not based on a folk song at all, but rather on a melody composed, more or less, by the Baroness de Béville. 

Baroness Adélaïde-Marie Yvelin de Béville

While I haven’t been able to dig up much about her, the Baroness indeed seems to have been rather a fascinating character. She was the author of a small pamphlet La vie de Marie, a collection of nine canticles intended, so the cover says, ‘to be sung in all the offices celebrated in honour of the most blessed Virgin’ (‘destinés à être chantés dans tous les offices célébrés en l’honneur de la très-sainte vierge’). They were published as an offering of ‘denier de Saint-Pierre’, a donation to the Holy See. You can read them online—thanks this time to the digitisation project at the Bibliothèque nationale de France—and they are very similar indeed to the text of La cloche sonne, leaving little doubt as to the veracity of Liszt’s inscription. The Baroness also wrote another pamphlet, a copy of which has so far eluded me, La charité de l’avenir (Paris and Brussels: Régis Ruffet, 1865), which is listed in the 1865 Journal général de l’imprimerie et de la librairie (Volume 9, Issue 1, page 65, for those of you who care). She pops up again, somewhat incongruously, in the Comptes rendus hebdomadaires des séances de l’académie des sciences, vol 63 (July-December 1866) (Paris: Gauthier-Villars, 1866), with a ‘communication relative au choléra’, specifically ‘the use of lemon as an anti-poison for the most subtle venoms and in particular for cholera’. 

Most interestingly, a somewhat weightier tome, Le Livre de toute la vie: Jésus, premier bonheur, dernier espoir (Paris: Malé, 1865) also furnished texts for a musical publication of her own composition, Cantiques du livre de toute la vie (Paris: Malé, 1866) for four-voice choir and organ, which publication was reissued by Ruffet around 1872.

Liszt was kind of awesome

Before getting into more of the musical nitty-gritty, I think this situation itself is worth noting for what it tells us about Liszt. Despite spending the 1860s touring extensively in Germany, Austria, Italy, France, Belgium, and even England and Russia, he was also engaged in 1861 as Director of Music at the Vatican, Pope Pius IX being among his many great admirers. In this role, which he held until 1865, he oversaw music—and, indeed, played the organ—for many services. His own religious faith is well-known, and his extensive output of choral works attests to the inspiration he drew from it. It was perhaps during one of his stays in Rome that he met the Baroness de Béville, although there is no reason that they didn’t meet when Liszt was residing in Paris in the 1830s and 1840s. According to the Sotheby’s listing, the later manuscript is accompanied by a letter from Rome dated 10th September 1866. In it, according to the lot description, Liszt writes of the inadequacy of music to capture certain emotions, and thanks the Baroness for her inspiring Angelus. 

Franz Liszt in 1870.

Looking again at the ‘melody page’ of the earlier manuscript, I am reminded of those times when one of my younger students had ‘composed’ something and needed my help writing it down. There is here a certain rough-and-readiness that I recognise, with corrections and alterations—in particular, the question-mark above the third bar makes me smile—which comes from taking dictation from someone not necessarily completely clear in intention, trying to fix in writing something that comes out slightly differently with each rendition. That Liszt—who was at this point as celebrated a musician as anyone had ever been, not to mention a pianist and composer of the most exceptional accomplishment—took the time, in a lifetime of frantic productivity, to be of help to the Baroness, and to do so with such good grace, professionalism, and charm, speaks volumes about his generosity and humility. That he turned her melody into such an effective musical miniature is yet further proof, if any were needed, of the incredible fertility of his creative mind.

The second manuscript gives the instrumentation ‘Harmonium (or Pianoforte)’, and the dating of the letter places this piece alongside other original harmonium works and arrangements for the harmonium that Liszt made towards the end of the 1860s. The harmonium version is now officially catalogued as S.663c, while the ‘piano’ version is S.238. It’s worth bearing in mind, of course, that the Weimar manuscript doesn’t actually specify an instrument; it was assumed that it was for the piano. So, while the later manuscript offers us the benefit of some performance directions, of which the Weimar version is completely devoid, the piano seems to me to offer a more appropriate sonority to the tolling effect of the repeated notes in this piece and, as I will discuss later, I think some departure from Liszt’s markings will enhance that image more fully.

But first, let’s look at the Baroness’s words and melody, because it gives us our interpretative bearings. The text, at least as much of it as has been preserved in Liszt’s manuscript, reads:

la cloche sonne l'angelus 
au loin dans la sombre campagne
la fleur tombe de la montagne, 
la fleur aussi dit l'angelus
la fleur vous couronne, O ma vie
Ave, ave Maria, ave, ave Maria

dans la nature l'ange prie
the bell tolls the Angelus 
far away in the dark countryside
the flower falls from the mountain,
the flower, too, says the Angelus
the flower crowns you, oh my life
Ave, ave Maria, ave, ave Maria

In nature the angel prays

(Before anyone pounces, the correct French orthography is angélus, I know; but my expertise is, alas, insufficient to say whether this was the case in the 1860s. In the later manuscript, the inscription does have several accented characters, but ‘angelus’ persists, so I’ll leave it as Liszt wrote it.)

What is the Angelus?

The Angelus is a Catholic prayer of Marian devotion, with origins as far back as the 11th century. A part of the daily liturgy, it is a Versicle and Response, interspersed with three recitations of the Hail Mary (the Ave Maria mentioned in the Baroness’s poem) and followed by a prayer. It is usually recited three times daily, at Matins, Prime and Vespers, and accompanied by the ringing of the Angelus bell, which serves as a calling of the faithful to prayer. The Angelus bell is traditionally rung in three groups of three peals. The text is as follows (the Hail Marys are omitted):

Angelus Domini nuntiavit Mariæ,
Et concepit de Spiritu Sancto.
Ecce ancilla Domini.
Fiat mihi secundum verbum tuum.
Et Verbum caro factum est.
Et habitavit in nobis.
Ora pro nobis, Sancta Dei Genitrix.
Ut digni efficiamur promissionibus Christi.
The Angel of the Lord declared unto Mary,
And she conceived of the Holy Spirit.
Behold the handmaid of the Lord.
Let it be done to me according to thy word.
And the Word was made flesh.
And dwelt among us.
Pray for us, Holy Mother of God.
That we may be made worthy of the Christ's promises.

The angel in question, of course, is Gabriel; the story, that of the Annunciation. So, the text paints a picture of the tolling Angelus bell, ringing across the dusky landscape, perhaps from the mountaintop, the rural setting, the countryside and the mountain, and the flower’s response to the call to prayer, all mix a Romantic pantheism into the historical imagery of the liturgical rite. 

Subtle tone-painting

The detail above is more than mere background because Liszt’s tone-painting is, I believe, quite precise. The introduction can be seen, and phrased, as three groups of three bell-peals, the traditional Angelus pattern, the first two peals of each three echoing back to us, something like this:

(More advanced players might experiment with half- or three-quarter-pedalling to make this even more effective.)

This pattern continues in the accompaniment until bar 11: chime-echo, chime-echo, chime. As the piece progresses, of course, Liszt relaxes this rhythmic scheme, which is exactly as it should be. The coherence of the piece as a whole should always take precedence over a strict adherence to something like musical imagery. The aim is to create an effective musical structure, not merely an accurate impersonation. The bell-effect does persist in the middle section, too, with the repeated dominant pedal note in the lowest voice.

Note also that, despite being in the key of C minor, there are only three B♮s in the entire 29 bars of the piece, and Liszt avoids A𝄬 entirely until bar 13. This gives the first section a modal flavour, which adds to the air of religiosity (and, indeed, also explains the initial assumption that the melody was a folk song).

The music

The piece can be seen either as an altered ’rounded binary’ structure, having two distinct sections and a codetta based on the A-section, or as a ‘truncated ternary’ form, with the A-section repeat halved in length and altered to provide harmonic closure. The A-section, beginning with a three-bar introduction, presents an eight-bar melody in triple time, characterised by a dotted rhythm on the upbeat. (Take care, by the way, not to bump when the phrase begins on the offbeat.) The compass of the A-section melody is contained within the perfect fifth, tonic to dominant; this allows a mutual reinforcement between the melody and accompaniment through shared overtones. Unusually for a piece at this level, the melody is entrusted to the left hand—another selling-point, if it were needed.

The B-section, bars 13-24, in duple time, presents a new melodic idea in the soprano voice, in the relative major. Alto and tenor voices in descending thirds fill in the V7-I progression, which descends over a tolling dominant pedal. The turn back to C minor at bar 20 ushers in a reshaping of the B-section melody, before the triple meter returns to close with an altered version of the A-section melody.

Notice also how the A-section opens and closes ambiguously: the absence of a third in the opening chord leaves the tonality uncertain until the melody enters, while bars 11-12 hover at the unison dominant and, rather than resolving cadentially, recontextualise the note G when the B-section begins in bar 13 as the mediant of the relative major, in which role it first serves as a sort of dominant 13th appoggiatura.

Liszt also carefully avoids any sense of cadential resolution: despite the key of C minor, it isn’t until bar 20 that we hear a leading-note for the first time. Significantly, this is the first ‘proper’ dominant chord, harmonising the exact dominant note that was left unresolved at the close of the first section, nicely framing the middle section.

This dominant chord is prolonged across the start of the new phrase, and while resolving subtly at bar 22, the strength of the perfect-cadence progression is sapped by the dominant seventh chord being in first inversion and the absence of the fifth from the tonic chord that follows it. This harmonic resolution is not the focus of the phrase in bars 21-25 in any case: that honour is bestowed upon the Neapolitan sixth in bar 23, one of only two chromatic alterations in the entire piece. It is a striking moment, proof that less is more—and that even a virtuoso as flamboyant as Liszt had the good musical sense to understand this. Although the Neapolitan is used in its traditional way, that is, as an altered supertonic preparation for a perfect cadence, that cadence itself (bars 24-25) is less than emphatic, landing on a first-inversion tonic chord, and the texture dissipating to only two voices.

One point to note is that there are several recordings that I’ve listened to—including the ‘official’ one accompanying the ABRSM exam syllabus—that play the B in bar 24 as B𝄬. It’s not. Both the Weimar and the Sotheby’s manuscripts have a natural sign in front of this note. Where this reading has come from, I’m not sure, but it is wrong. Speaking with my editor’s hat on, even if the natural sign were missing from either or both sources, I would readily suggest it on the grounds that without it the harmonic progression is much less effective; especially following the Neapolitan in the previous bar, it is really the only reading that makes sense.

It is only the re-entry of the first-section melody in the bass that establishes the root-position tonic harmony. The final five bars function as both a codetta and a truncated recapitulation of the first-section. Outlining the progression i-ivc-ic-ivc-i, the harmony here feels not only grounded but, indeed, almost static owing to the tonic pedal beneath the final plagal cadence. The F♯ in bar 27 introduces a little Hungarian seasoning, the augmented fourth deriving from traditional gypsy scales (such as those in the B-minor sonata). Again, Liszt’s refined musical judgement is evident in his varying the repeat just enough to avoid monotony.

The codetta closes with the right-hand chord both providing triadic completion of the opening fifth, an octave lower than where it began. Just as the B-section is framed with the octave Gs, first alone then harmonised, the piece itself comes full circle in the same way.

This is a gem of a piece, ingeniously constructed, with vivid musical imagery and a real sense of atmosphere. I do hope you enjoy it!


Transcriptions of the sources and a performance edition are all available in the download library.

I am grateful to Adam Thomas for his help with musical typesetting.

Cécile Chaminade: ‘Automne’, op. 35 no. 2

Undoubtedly one of Chaminade’s most well-known and well-loved solo piano works, ‘Automne’ comes from the collection Études de Concert, op. 35. No fewer than four of the six etudes that comprise the set feature on diploma works lists. They are all gratefully written, and although they deserve the ‘etude’ title, their musical interest and audience appeal also fully justify the time invested in studying them. Chaminade herself seems to have held ‘Automne’ in particular esteem: Citron lists five public performances of this etude, including in London, Paris, and Philadelphia (which is more than her performances of the other five etudes from opus 36 combined).

As an etude, the primary technical focus in the A-section is on voicing and phrasing, while the B-section requires good rotation, chord-playing, and a good sense of keyboard geography. Particularly careful attention to the pedalling will also help, and there is an abundance of tempo-markings to make sense of. Although it’s not without challenges, ‘Automne’ is a very effective concert piece when played with technical assurance and musical authority. Let’s dive in!


Analytical commentary

The work is in ternary form, with an almost exact repeat of the A-section. The A-section itself is a ternary form, while the B-section approximates a rondo shape. The A-section’s atmosphere is redolent of Keats’s ‘mists and mellow fruitfulness’, with its lyrical melody and warm harmonic accompaniment, while the B-section is rather stormier.

Aabb. 1-7
bbb. 8-17
a’bb. 18-29
Babb. 30-37
bbb.38-49
abb. 50-59
cbb. 60-71
a’bb. 72-77
Aabb. 78-84
bbb. 85-94
a’bb. 95-107
Chaminade: Automne, op. 35 no. 2: summary structural breakdown

The construction of the A-section is somewhat typical of Chaminade’s structural approach. The six-bar opening paragraph consists of an antecedent-consequent phrase-pair, followed by the antecedent phrase again, altered to move to the dominant:

Chaminade: Automne, op. 35 no. 2: bb. 1-7

Melodically, the opening rise outlines an octave between the upper and lower dominant, and both subsequent phrases are contained within this compass. In contrast, the second phrase (beginning b. 8) breaks out of this constraint, pushing chromatically upwards to B♭, then sequentially to E♭. Harmonically, the music shifts to the minor mode of the dominant, A♭, which is established temporarily as the tonic by the end of bar 7. This in turn functions as the minor subdominant of our destination in the new phrase, E♭ minor, which arrives at bar 9. The process is then repeated taking us to A♭ minor at bar 11. Both phrases confirm the arrival of these keys by means of a ib–V7c–i progression, the bass falling stepwise through mediant and supertonic to the tonic. This second paragraph shares a rhythmic outline with the previous one, although the chromatic push upwards is now balanced by three repeated notes, rather than the ascending broken chord of the parallel phrase in the first paragraph.

Chaminade: Automne, op. 35 no. 2: bb. 8-10

Bars 12-15 comprise another sequential passage which takes us to a restatement of the opening subject in the bass, in B♭ major. This leads, via a diminished seventh, to a perfect cadence back into the home tonic at bar 18 for a repeat of the a-section, texturally expanded, and extended with a codetta (bars 24-29). 

Although this A-section is an ostensibly ternary structure, it is worth observing that the melodic relationship between the a- and b-sections obscures these structural divisions and, along with the irregular phrase-lengths, creates instead a feeling of ongoing development. The ternary structure is felt most strongly in the texture. The a-section is homophonic, even if the melody avoids the soprano until the restatement in bars 18-23; see also how the codetta returns to the texture and tessitura of the opening. The b-section is more imitative, with the soprano and bass often engaging in contrary motion, and the accompanying chords lying in between; even in bars 16-17, where the bass melody appears to dominate, the soprano soon pushes us away from the B♭-major territory with an ascending chromatic conjunct minor third, F to A♭, thereby steering us back to A♭ major.

(There is a wonderful textural touch at bars 20-21, by the way, where the tenor joins the soprano to double the melody an octave lower.)

This content is for members only.
To continue please Login or Join Now

Edvard Grieg: ‘Elf-Dance’, no. 5 from Lyric Pieces, op. 12

Inspired by the folklore of his native Norway, Grieg’s ‘Elf-Dance’ is a scherzo-like movement which exemplifies many aspects of Grieg’s early piano writing. Most often female, elves in Norse legend are mainly to be encountered at night or in the mists of early morning, dancing in groups, leaving behind ‘elf-circles’. Frequently associated with illness or bad luck, should a human happen to observe their dance, while it feels to them that only a few minutes have passed, in reality many years have gone by. In this depiction of one such dance, there is indeed both a ‘circularity’ and a hint of menace. 

Like most of Grieg’s Lyric Pieces, ‘Elf-Dance’ eschews the developmental principles associated with Germanic sonata-form, consisting of paragraphs that are repeated and varied in pitch or in details of melodic contour. (This manner of construction, too, can be seen as a result of Grieg’s interest in Scandinavian folk-music.) Much of the content of these larger paragraphs, it will be seen, are in fact drawn from a small number of ideas presented in the first four bars. 

Richard Doyle (1824-83): Fairy Rings and Toadstools (1875), watercolour on paper.

Pianistically speaking, the piece is idiomatic, effectively exploiting contrasts of register and tone-colour to enhance the imagery suggested by the title. The figuration is gratefully placed, allowing the performer to achieve the requisite delicacy of touch without any technical awkwardness. This is perhaps unsurprising, given that the pieces comprising what was to become the first of Grieg’s ten volumes of Lyric Pieces began life as teaching pieces for the composer’s own use. 

Analytical commentary

The scherzo-like ‘Elf-Dance’  is in E minor, and remains within this tonal orbit throughout, relying heavily on diminished chords to create an atmosphere of mystery and danger. The form might be described as a binary-rondo hybrid:

Abars 1-8
Abars 9-16bars 1-8, repeated exactly
Bbars 17-30transformation of rhythmic and melodic motifs from A section
Abars 31-38bars 1-8, repeated exactly
Bbars 39-52bars 17-30, repeated exactly
Abars 53-60bars 1-8, repeated exactly
Codabars 61-72based on A

However, both A and B sections rely heavily on a limited number of motifs introduced in the first four bars, which are either varied or, as it were, superimposed; that is, the rhythmic profile of one motif is combined with the melodic contour of another. While this shared material creates a feeling of a unified whole, the sectional divisions outlined above represent paragraphs of contrasting mood and harmonic emphasis.

Ex. 1: Grieg, ‘Elf-Dance’, op. 12 no. 5, bars 1-3

The A section presents two ideas, whose rhythmic figures also feature in the B section. The first, a five-crotchet pattern, consists of repeated tonic chords separated at the midpoint by one submediant chord. The second is a delicately ascending quaver figure, accompanied below by a similar chordal pattern, now shorn of its final crotchet, featuring an anacrustic conjunct third which is also to feature prominently in the B section (see ex. 1). Harmonically, the movement in this second phrase is from tonic to subdominant. The melody having ascended to the dominant at the end of bar 4, bars 5-6 repeat bars 1-2, transposed upwards into the dominant minor. For the concluding phrase of the paragraph, the left hand takes the melody, which again features the conjunct rising third figure (see ex. 2). At bar 8, the harmony moves into the dominant major; this open-ended progression opens up both the repeat of the A section (bars 9-16), and the beginning of the B section.

Ex. 2: Grieg, ‘Elf-Dance’, op. 12 no. 5, bars 7-8

The B section represents the more menacing aspect of the music’s mythical subject by means of more widely contrasting dynamics, a shift to a lower register of the keyboard, and the harmonic ambiguity of diminished-seventh chords, which feature prominently throughout this section. This ambiguity extends, in fact, to the key of the entire section. The continuation in the dominant area is implicit from the melodic emphasis upon Bs and F-sharps; but there is no suggestion of major or minor tonality from bar 17 until the preparation of the perfect cadence in the global tonic in bar 30.

The melodic descending perfect fifth from bar 8, F-sharp to B, is inverted for the opening phrase of this section. The phrase, the hands playing an octave apart, borrows its outline from the opening five-chord motif, repeated Bs now divided at the midpoint by F-sharp. The oscillating quaver accompaniment from the right hand in bars 7-8 is now combined with the anacrustic conjunct third. The same rising third is presented in augmentation in the bass, again borrowing the five-crotchet rhythmic figure (bars 19-20), while the lower notes of the right-hand oscillating quavers reverse the pattern, in contrary motion with the bass, again simulating the ‘circularity’ of the elves’ dance (see ex. 3).

Ex. 3: Grieg, ‘Elf-Dance’, op. 12 no. 5, bars 17-20

Following an exact repeat of this phrase (bars 21-24), the contrary-motion pattern is repeated, transposed up a minor third (the construction of the diminished seventh chord, of course, means that while the musical material has ascended in pitch, the harmony remains uniform). One further repeat (bars 27-8) is compressed, the anacrustic conjunct minor third now a major second filled in chromatically, again falling but now joined by the left hand in similar motion, taking the music through three relatively static bars (bars 28-30), the harmony moving through a French sixth, to a dominant chord, the suspended fourth in the top voice resolving to the leading-note and then seamlessly connecting to the next statement of the A section at bar 31. 

From this point until bar 61, there is no change in the material. The brief coda evolves naturally from the first A-section phrase, taking the ascending figure, with its characteristic semitonal appoggiatura, through subdominant, Neapolitan, and dominant-seventh harmony, before one final reworking of the opening chordal idea. The five-chord figure is extended by two chords, indeed two statements of the idea are telescoped together, the final two chords of the first assuming the role of the first two chords of the second. This extended phrase affords the opportunity for the inner voices to move apart, then together, on the third and sixth chords of the phrase respectively—one final circular image before the final tonic chords ascend into the upper register.

Although bars 31-60 repeat bars 1-30 literally, there is a slight variation in the order in which the material is presented. The two statements of the A section, at first presented in succession, now frame the B section. As the coda begins with the first four-bar phrase of the A section, it might seem that this is unnecessary as the three-bar cadential pattern with which the B section ends would have prepared for the coda, in terms of texture and harmony, just as effectively. However, given the coda’s brevity, the preceding eight bars balance its mood; without this preparation, the coda would seem rather too brief and perfunctory. It is as if, enchanted once more by the magic of the dance, the baleful mood of the B section is forgotten, before the dancers disappear from sight along with the final glimmers of moonlight before the dawn.

Technical suggestions

This content is for members only.
To continue please Login or Join Now

Cécile Chaminade: Piano Sonata, op. 21, second movement

We begin our survey of Chaminade’s diploma-level piano works with the middle movement of her only piano sonata. Published in 1895, it is dedicated to Moritz Moszkowski, Chaminade’s brother-in-law. Marcia J. Citron cites a performance of the first movement given two years before the work was published by the composer herself in St. James’s Hall, London, on June 1st, 1893, while the third movement is the same as Appassionato, op. 35 no. 4, which was published with the other Études de concert in 1886. The earlier opus number implies composition in the late 1870s or early 1880s, as other works from that time were also not published until the 1890s or later. (Her Pastorale enfantine, op. 12, for instance, was published in 1897, and her ‘fantaisie très facile pour un seul piano à 8 mains’, Les noces d’argent, also bizarrely given the opus number 12 was published by Enoch in 1892, and a version for piano solo in 1907.) At the early stages in her compositional career, it is perhaps understandable that she should want to present only her best and most recent work for publication, and thus public scrutiny, and only later revisit juvenilia when her public and professional reputation was more firmly established. Still, the wholescale repurposing of the final movement of a piano sonata, a significant and serious compositional endeavour, perhaps suggests that, at least by the mid-1880s, Chaminade had no intention of publishing it. It may be that a favorable response to her performance (or possibly performances) of the first movement in the early 1890s changed her mind.

Although there doesn’t appear to be any record of her playing the second and third movements in public, Chaminade’s public performance of the first movement alone, and her re-use of the finale, implies that she was comfortable with the practice of excerpting sonata movements, so we should feel no qualms about giving the other movements the same treatment.

This is a beguiling movement, which certainly repays the modest effort required to learn it. Like most of Chaminade’s so-called ‘salon pieces’, it is at least ostensibly in ternary form. The A-section is in A♭—in terms of the sonata to which it belongs, the submediant (a pattern it shares with two of Beethoven’s sonatas in the same key, op. 10 no. 1 and op. 13). The B-section is in B major (enharmonically the flat mediant of A♭) and although the transition into and out of the B-section is effected by the same simple means (a chromatic figure in sixths), A- and B-sections share some thematic material, displaying an awareness of the high-minded nature of the nineteenth-century piano sonata. Stylistically there is a wonderfully nostalgic atmosphere; the chromatically-tinged harmony and melody-dominated homophony in the outer sections remind me of Ireland, or earlier Bridge, while the use of a restricted rhythmic vocabulary and the more imitative texture of the B-section seem reminiscent of Schumann.


Analytical commentary

The 118-bar movement falls broadly into ternary form, whose boundaries are defined by a modulation from the conventional dominant area to the flattened mediant, and then back again. (Interestingly, this local modulation from the dominant, E♭, to its flat submediant, C♭ (written enharmonically as B major), reflects the key-relationship between this movement and those which surround it in the sonata.

Abb. 1-39A♭ major36bb + 3bb bridge
bb. 1-7A♭ majorA1
bb. 8-15C minorB
bb. 16-24A♭ majorA2bb. 16-19 = bb. 1-4bb. 20-22, sequential extension + 2bb. codetta
bb. 25-36C5bb. + 6bb.
bb. 37-39E♭ majorbridgebased on main rhythmic motif, chromatic inflection
Bbb. 40-83(C♭=) B major
bb. 40-51B majorD1Two repeated six-bar phrases,bb. 40-45 moving towards D♯ minor,bb. 46-51 moving to half close in F♯ minor
bb. 52-56F♯ majorCFirst (5bb.) phrase
bb. 57-67B majorD2repeat of bb. 40-51, slightly elaboratediv-Ic-V in F♯ minor (b. 51) omitted
bb. 68-74B majorEfree sequential material2+2+3 bb. over dominant pedalfinal phrase recalling idea of D1
bb. 75-80B majorCSecond (6bb.) phrase
bb. 81-83E♭ majorbridgebb. 37-39
Abb. 84-106A♭ majorrepeat of bb. 1-23
CODAbb. 107-118A♭ majorbased on B, 12bb. tonic pedal
Structural breakdown of 2nd movement of Chaminade: Piano Sonata, op. 21

The first 23-bar paragraph divides into a ternary form of its own. (This structural procedure is typical of Chaminade’s works—see, for instance, the First Arabesque, op. 60.)

The first four bars (see ex. 1) present the material from which most of the A-section is derived. A dotted-rhythm motif, essentially a diminished-seventh chord on the tonic, momentarily decorated by upper and lower auxiliary notes in the soprano and bass, is followed by two ascending fourths which are harmonised IVb-V7-I6. This opening phrase is answered by a phrase built on rising and falling fifths. Note the rhythmic unity of the entire phrase: just two motifs persist throughout.

ex. 1: Chaminade, op. 21, second movement, opening

Bars 5-7 dovetail the close of the opening phrase and a chromatic ascending sequence (also featuring the rising fourth) which terminates on another diminished seventh, this time on the dominant.

Bars 8-15 make a B-section in C minor (see ex. 2), which nonetheless initially gravitates towards the subdominant of that key; until the assertive drone bass, which appears in b. 12, confirms the key, as do repeated statements of the tonic chord. Note the use of the opening rhythmic pattern—but now displaced by a beat, making the dotted rhythm an anacrusis—with a very different expressive quality. The retransition to the A-section material—simply omitting the dominant of the C-minor chord and reharmonising the E♭ and C with A♭—is, admittedly, a tad clunky.

ex. 2: Chaminade, op. 21, second movement, bars 8-16

Bars 16-19 are a literal repeat of bars 1-4, with just a slight alteration at the end of bar 19 to keep the music in the tonic for a three-bar sequence, heavy on plagal seasoning, closing via a vi-ii7-IV/V-V7-I progression.

Bars 25-36 consists of a 5-bar phrase, which is repeated and extended by a bar and taken in a new harmonic direction. This passage develops the opening dotted-rhythm motif, in a new harmonic context and a higher register and with a lighter texture (see ex. 3). Where the opening paragraph emphasised ambiguity and instability—with diminished-seventh chords and chromatic part-writing—this section revels in the tonic, established by means of repeated imperfect and plagal cadences punctuated by rising fourth-based chords over a tonic pedal.

ex. 3: Chaminade op. 21, second movement, bars 25-7

After returning to the register and voicing of the opening, the music modulates to the dominant, and a three-bar codetta recalls the dotted-rhythm motif over a chromatic rising and falling idea in sixths (see ex. 4).

ex. 4: Chaminade op. 21, second movement, bars 37-9

The B-section is rather Schumannesque in terms of its imitative texture and melodic lyricism (see ex. 5). It is notable for avoidance of the tonic, which features only occasionally in harmonic progressions moving elsewhere, and rarely in root position. As the duet between soprano and tenor develops the music is pushed towards D♯ minor, where it briefly settles at bar 45 before the process begins again, turning through parallel minor to a half-close in F♯ minor.

ex. 5: Chaminade, op. 21, second movement, bars 40-5

In a constructional masterstroke, the material from the A-section’s C-section returns, now in dominant major. This brings the A- and B-sections into alignment, in that both begin with passages of more turbulent, searching chromatic harmony, which are then followed by this almost serene passage of radiantly stable triadic harmony. This version of the paragraph is truncated, however, ending after four bars with a perfect cadence in the dominant.

This immediately gives way to a slightly altered repeat of bars 40-50 (bars 57-67), the music this time avoiding the final turn to F♯ minor, F♯ functioning as a dominant pedal throughout a climactic seven-bar phrase which subsides with one final reminiscent soprano-tenor exchange of the B-section’s principal melodic idea before the C-theme returns in the local tonic. The section concludes with a three-bar bridge passage—B, now enharmonically becoming the flattened submediant of E♭—based again on the chromatic-scalic idea in sixths.

The reprise of the A-section (bars 84-106) is, the slightly different opening aside, a literal repeat of bars 1-23. The B-section theme interrupts proceedings at b. 107 (see ex. 6), however, the harmony here oscillating between I7 and IVc. The tonic pedal remains until one final perfect cadence at bars 116-17 brings the scene to a calm conclusion.

ex. 6: Chaminade, op. 21, second movement, bars 106-109

I hope this brief discussion shows that, although it is not inaccurate, to describe this movement as a simple ternary form does not do justice to the relationship between the material and the tonal shape of the movement. The rhythmic unity, as well as the ‘through-composed’ sense, of this movement shows a more sophisticated approach to form than many of Chaminade’s single-movement miniatures. Given the thoughtful structural integration demonstrated in this early essay, it is to be regretted that the composer never made subsequent forays into the more ‘serious’ world of the piano sonata.

Performance Issues

There is little here that a student should find overly taxing. Well-voiced chord playing, and careful legato-pedalling are required, along with sensitivity to the B-section’s imitative texture. Musical interest will derive from imaginative ‘orchestration’ and varied tone-colour. Don’t shy away from use of the una corda pedal, and not just for the softest passages. The contact between the less-densely compacted hammer surfaces and the strings results in a gentler attack, not just a quieter dynamic—just as a muted brass or string instrument can play at varying dynamics—so, too, the una corda offers us a different tonal palette at a range of tonal levels.

The C-theme’s right-hand arpeggios can be taken consistently with a 1-2-3-4-1-2-3-4-5 fingering (see ex. 7). I would advise against adopting varying fingerings for each chord; although graded-exam fingering patterns are often designed to provide ergonomic solutions to the specific problems posed by scales and arpeggios in the various keys, it is a fact of life that advanced pianists must be comfortable playing with our outside digits on black keys. In this case, in the second set of arpeggios the fourth finger will land on Bs, and second and fourth fingers will land on Bs and Es respectively in the final iteration. With a supple hand and wrist I believe the consistency will outweigh these mechanical inconveniences.

ex. 7

A comfortable octave span is essential—there are some tenths, but most can be spread, even when not marked, or even redistributed. There is, for example, no reason for the tenor A♯ in the first chord in bar 56 not to be taken by the right hand, nor the tenor B in the first chord in bar 58. Likewise the last two alto notes in bar 69 can be taken with the left hand; the leap from the octave F♯s in the bass is fairly acrobatic, but perhaps still preferable to breaking the right-hand stretch. This is not an exhaustive list—but bear in mind that Chaminade doesn’t use the treble- and bass-clef staves to allocate notes to the right and left hands respectively. Rather, the hands move freely between staves in preference to frequent changes of clef. Bars 9-12 require clef-changes in the lower stave—there is no better alternative in this circumstance—but by and large the staves retain their ‘standard’ cleffing and the hands move between them. (The give-away is empty staves without rests; the hand in question must be doing something on the other stave!) In performance it is more important to adhere to the spirit of the musical product than to follow the letter of its notation, so feel free to explore creative solutions to the more widely-spaced chords. (Bar 107 offers a concrete example of one such solution, which could be applied elsewhere)

An instance such at the left-hand acciaccatura in bar 108 would seem to imply, by the absence of the same device elsewhere, that if you plan to use a similar approach to chords that are too wide to be taken in one go, the lowest note should coincide with the right hand, rather than being placed before everything else. There is little reason for the notation in bar 108 unless it is an exceptional circumstance. This, obviously, would not apply to ‘harped’ chords across the entire grand stave, but in the case where the arpeggio is restricted to one stave or the other, again, perhaps the bottom note of the spread chord should be placed on the beat.

I would recommend a considered approach to tempo. This is a danger area as inconsistency will result in a rendition that sounds diffuse. The overall Andante should not, in my opinion, be sluggish—andante means ‘going’!—and although the B-section is more flowing, with more regular phrasing and more consistent quaver motion, it is not given a different tempo marking, so the performer must find a ‘baseline’ tempo at which both sections work effectively. The frequent animato instructions should also be moderated. Every one of them coincides with a crescendo, so if an acceleration is required to convey a more animated feeling, I would reserve it until the peak of the crescendo. None of the animato markings is followed immediately by an a tempo, which only occur after poco rit., rit., or rall., so there is nothing to imply that Chaminade intended animato specifically to mean ‘faster’. Each of the ‘slower’ markings, on the other hand, coincides with a diminuendo, so the player must decide: do these markings convey a drop below baseline tempo, and the a tempo markings a return to it? Or do animato+crescendo and rit.+diminuendo balance each other, the a tempo confirming the resumption of the baseline tempo?

Note the consistent distinction between rit. and rall. markings. Rallentando can only be interpreted as a gradual deceleration, but rit.—an abbreviation of either ritenuto (immediately slower) or ritardando (synonymous with rallentando)—could be either. There is only one rall. here (in bar 71), but its presence implies that Chaminade intends the various rit. markings to mean ritenuto, not ritardando.

Tellingly, the final page indicates no change of tempo, and I think this is the interpretative clue to the tempo puzzle. No performance of music from any era (some twentieth-century music being an exception) should be rigid, because it lacks breath and, ultimately, that breath is what gives the performance humanity. But consistency is not rigidity; a subtle capacity for flexibility in phrasing, a sense of musical punctuation, is just as vital a component of musicianship as a solid sense of pulse. This is not a license to be wayward, however, as this waywardness turns charm and sensitivity into mannerism and sentimentality. These are charges often levelled at composers of so-called ‘salon music’ (in which category much of Chaminade’s output resides). Often, however, the guilty party is not the composer, but rather the overly indulgent performer. To do justice to the economy and integrity of the compositional structure, aim for a performance that is similarly organised and intelligent.

Verified by MonsterInsights