I love art songs, and I've recently returned to practicing art song accompaniments on the piano. It's been good for me; I get to practice music I like without the usual piano exercises or repertoire. Moreover, I get to rethink, reconsider, and reaffirm my philosophy regarding piano playing—especially as it pertains to the accompaniment of song. Although I may contradict much of what is accepted piano playing philosophy, this is, nonetheless, how I feel this music should be played. 

Currently, songs on which I've been working are: 

1. Franz Schubert: Erlkönig, Gretchen am Spinnrade, Die Forelle, Stänchen, Das Wandern 

2. Gabriel Fauré: Notre Amour, Clair de Lune, Aurore, Les Berceaux, and the cycle Mirages, Op. 113 (rarely performed) 

3. Malcolm Williamson: from the cycle "A Child's Garden," The Flowers, My Bed Is a Boat, A Good Boy 

4. Roger Quilter: Fair House of Joy, Love's Philosophy (phew!) 

5. Ralph Fisher: Somewhere i have never travelled, is there a flower, if i have made, my lady, a clown's smirk (hard) 

6. Ralph Vaughan Williams: Whither Must I Wander, The Vagabond 

Then there are those wonderful women composers (predominately American, mostly with three names) of the early part of the twentieth century: 

1. Mildred Lund Tyson: Sea Moods, The Lilacs Are in Bloom 

2. Mary Turner Salter: The Cry of Rachel (Did someone say melodrama?) 

3. Elinor Remick Warren: Snow Towards Evening 

4. Teresa del Riego (British): Homing (guaranteed not to leave a dry eye in the house—when sung correctly, of course) 

5. Clara Edwards: Into the Night 

The thing that I have tried to achieve with these songs, as I practice them, is to articulate the notes and phrasing as indicated by the composer; something that frequently requires abstention from use of the sustain, or damper, pedal to a large extent. This is especially true in the Schubert and Fauré songs, but is also true in many of the others. 

Keyboard musicians are notorious for ignoring phrasing. Organists just get so wrapped up in their legato-obsession that they simply can't resist slithering around the instrument without a single phrase break in an entire piece whether it's indicated or (as in Bach) not. Pianists are just as guilty with their neurotic use of the damper pedal, often resulting in the piece being one continual blurry wad of sound. In short, keyboard musicians don't breathe. Unlike singers and wind players, they aren't physically compelled to do so, so they don't. Since most organists are also trained pianists (as opposed to the converse), and are often called upon to accompany at the piano, and, since I'm focusing on song accompaniment, we'll stick to the piano. 

Let's choose a Schubert song; say, Gretchen am Spinnrade. The first four measures tell us a lot (see Excerpt 1). 

Excerpt 1: Gretchen am Spinnrade, D. 118, by Franz Schubert, mm. 1-4.


First of all, the composer gives us no pedaling indications. This is not unusual since the sustain pedal (or lever) on the pianos in Schubert's and Beethoven's time was still quite rudimentary. Its use was limited primarily to special effects or to aid in achieving some semblance of legato for large stretches or other awkward sections. As a result of the lack of pedal markings, we must rely upon the pianist's intelligence and musical acumen. At the onset, the accompanist sees the terms sempre legato in the right hand and sempre staccato in the left hand. Typically, a pianist will more than likely focus on achieving a legato in the right  hand, since that's the faster moving part and symbolizes the motion of the wheel itself. Therefore, a discreet but steady use of the damper pedal will invariably ensue. Also, since there aren't any of those pesky slurs, the pianist doesn't have to worry too much about phrase articulation. Notwithstanding, a good pianist will at least try to give some kind of illusion of a detached left hand. But wait a minute! The left hand is in essentially two parts, an upper part consisting of one or more detached eighth notes and a slow moving, almost organ-like, single note dotted-half note bass line. Moreover, Schubert's sempre staccato indication appears below the staff. Now, does that mean ALL the voices in the left hand are to be played staccato? One would hope not. Obviously the upper part is to be played staccato, as indicated, and the bottom voice legato. Consequently, this totally negates use of the damper pedal. One could nearly achieve—perhaps—the desired effect of staccato and legato in the same hand with employment of the sostenuto pedal. But, why would anyone want to do that since there aren't any particularly difficult stretches in the left hand which would require its singular benefit? The sostenuto pedal is a handy-dandy device, and I love using it; but, it's unnecessary here. 

Of course, this does not preclude the sustain pedal altogether; rather, its discriminating application can often times be very effective. For instance, in the same song, Schubert briefly changes the mood to an almost dreamlike feeling when Gretchen begins to describe the man she loves—"Sein höher Gang, sein' edle Gestalt" (see Excerpt 2).

Excerpt 2: Gretchen am Spinnrade, D. 118, by Franz Schubert, mm. 51-54.


Here the composer has shifted to F Major, and the piano becomes much softer (pp); furthermore, while the right hand continues to keep the wheel spinning, the left hand has become static, as if she has lost track of her work completely and has begun to fantasize about him. Here the pianist can virtually sit on the damper pedal, reinforcing the trance-like state into which she has wandered. The effect can be quite startling in contrast to the very clear, almost contrapuntal preceding section. However, the atmosphere soon changes. The harmonic rhythm quickens, the dynamics increase, the tempo accelerates as she becomes more and more fervent in her fantasy, climaxing to the point of a passionate kiss (see Excerpt 3) 

Excerpt 3: Gretchen am Spinnrade, D. 118, by Franz Schubert, mm. 65-68.


I argue that, contrary to normal practice, the gradual lessening of damper pedal application actually increases the tension, creating a tautness and starkness to her fevered delusion. T

This also holds true for the second half of the song. After she gets back to work, she again goes into an emotional fervor—"Mein Busen drängt sich nach ihm hin." This time she yearns for her lover in even greater despair of his absence. Using the same techniques as before, Schubert builds to a desperate, one might say erotic, climax; after which Gretchen virtually collapses, or simply gives up (depending upon how far you want to take this), as the spinning wheel gradually comes to a stop. Again, I highly recommend eschewing application of the damper pedal. A held note with only its one damper raised (i.e., without the pedal) has a shorter sustain than if it's struck with the pedal down and the whole row of dampers raised. The result is that the sforzati in the left hand at the peak of both climaxes becomes much more pronounced and effective. Therefore, greater drama is achieved (see Excerpt 4).

Excerpt 4: Gretchen am Spinnrade, D. 118, by Franz Schubert, mm. 104-106.


Okay, so we know about the shortcomings of Schubert's piano and that his songs require less pedaling. What about Schumann, Brahms, Löwe, Wolf, etc.? Even with these later composers, although there are possibly more occasions for pedal usage, coupled with their often imprecise pedal markings, it is still quite evident that damper pedal usage can and should be reduced drastically—even avoided— provided that the pianist actually plays legato. 

And what about the French? Surely, French melodie, with its luxuriant harmonic language and elegant textures, lends itself to a more liberal approach to the sustain pedal. Au contraire, mon ami. This is a grave misunderstanding of the nature of French music, perpetrated by what I consider to be a general misreading of those composers who are commonly referred to as the Impressionist School, of whom Debussy and Ravel are supposed to be its chief exponents. With the excessive use of the sustain pedal (especially with those two composers), one very crucial aspect of the music—and the French musical sensibility in general—is overlooked or ignored: the love of clarity first and foremost. There is probably no greater exponent of this approach to composition than Gabriel Fauré. Fauré epitomizes all those qualities we admire in French song: transparency of textures, economy of means, elegant flowing melodies, eminently singable vocal lines, and perfect balance between the voice and the piano. Of course the tendency for pianists to treat Fauré's melodies (as well as most French song) as "Impressionistic" (i.e., with lots of pedal) is very shortsighted. The one thing I've noticed about Fauré's accompaniments is that he is very specific as to where the pedal is to be used; and that when it is to be employed—if at all—it's to be done so sparingly. 

Let's look at Clair de lune, as an example. This song is unique because the piano not only sets the mood, but is the actual focus of interest because of its melodic and harmonic consistency; whereas the voice, in an almost through-composed style, merely comments on this enchanting phenomenon portrayed by the piano; thus almost giving the impression of a piano solo with vocal accompaniment. For the first seventeen measures there is not one pedaling indication! Now, does that mean the sustain pedal is not to be used at all during those measures? In a word, yes. During this first part, the right and left hands are phrased separately and, therefore, should be articulated accordingly. The left hand is made of small groups of three sixteenth notes separated by a sixteenth rest at the beginning of each beat, each group with its own slur (see Excerpt 5). 

Excerpt 5: Clair de lune (Menuet), Op. 46, No. 2, by Gabriel Fauré, mm. 1-3.


The notes within these groupings are, of course, to be played legato. The right hand consists of a separate flowing melody of one- and two-measure phrases which are to be articulated distinctly from the left hand. The problem here is that most pianists simply refuse to acknowledge any phrasing; and their insistent use of the damper pedal simply exacerbates the problem. When good, fluid, legato technique is used there is no need to employ the sustain pedal in those first seventeen measures. My only caveat would be at mm. 5 and 13 (and duplicate passages), where the slightest tap of the pedal may be used to achieve legato between the repeated Fs (see Excerpt 6). 

Excerpt 6: Clair de lune (Menuet), Op. 46, No. 2, by Gabriel Fauré, mm. 5-6, pedal markings added.


This would be a legitimate function of the sustain pedal; but even then, I'm not fully convinced of the necessity. 

In any event, at m. 18 ("Jouant du luth et dansant") we finally see actual pedaling indications from the composer (see Excerpt 7).

Excerpt 7: Clair de lune (Menuet), Op. 46, No. 2, by Gabriel Fauré, mm. 17-18.

Here, they occur on the off beats and are only long enough to cover the value of the eighth notes in the left hand; at mm. 24 and 25, only the little arpeggios in the left hand are pedaled. The return to the opening melody in the piano at m. 26 ("Tout en chantant, sur le mode mineur") signals a return to the absence of pedaling. 

Then, at m. 38 ("au calme clair de lune, triste et beau"), with his very specific use of the damper pedal, Fauré transports us into a dreamlike world for four measures—shifting to a major tonality and full measure arpeggios and lifting the pedal only when the harmony changes (see Excerpt 8). 

Excerpt 8: Clair de lune (Menuet), Op. 46, No. 2, by Gabriel Fauré, mm. 38-41.


He does this again for two more measures at m. 44. The next pedaling occurs at m. 51 ("Les grands jets d'eau sveltes parmi les marbres!"), when he reprises the off beat eighth-note-length pedaling of earlier; this time across three measures of a repeated seven-note pattern under one phrase marking. 

One last set of pedal indications occurs four measures from the end (mm. 58 and 59), on the G-flat major and B-flat minor triads, with no pedaling in the final two measures (see Excerpt 9).

Excerpt 9: Clair de lune (Menuet), Op. 46, No. 2, by Gabriel Fauré, mm. 58-61.

As we see, Fauré is very specific as to where the damper pedal is to be applied. The result is that when the pedal is employed, the unique effect it has on the mood of the song is greatly enhanced. This effect would be lost if the pedal were used regularly throughout, even if applied carefully. 

One factor to keep in mind is that many French composers of Fauré's generation and earlier were organists—including Fauré himself. So, legato touch sans pedale on the piano is not a foreign idea for French music. Even composers who were not necessarily organists (e. g., Debussy and Ravel) were either influenced by their organist predecessors and contemporaries, or simply preferred the clarity and subtle nuance that can be achieved by minimizing the use of, or simply avoiding, the damper pedal. 

Improvements over the years in damper pedal technology have made it easy for pianists to abandon true legato playing. Consequently, I am presenting the pianist with two challenges: the first is to learn to play legato. That means holding down the note for its full value before before playing the next note, and often employing the art of finger substitution. This is, in simple fact, technically much more challenging than using the damper pedal for legato, making what might have been a modest technical matter become a rather formidable task. 

Secondly, as part of this "pedal-less" approach, the pianist must learn to respect the phrase! Composers don't put those slurs into their scores just to clutter up the page. The idea is that, at the end of one those curvy lines, one is supposed to breathe. Just ask any wind player or singer. The fact that the piano's phrases may be longer or don't directly coincide with the vocal part is the whole idea! You have these two or more individual conceits, each with its own integrity, so to speak, coming together producing an integrated whole. That doesn't really happen if the piano is only in the background, merely supplying "mood" for the song. I

chose these two songs because they are vastly different songs stylistically and chronologically, and yet they require very similar approaches to accompaniment. Most of what I have discussed here can and should be applied to many, if not most other, art songs. The challenge arises in developing a sense as to where and when discriminating use of the damper can be exercised. 

So, what's to be gained from all this non-pedal approach to accompanying art song? First is clarity of line; instead of a vocal part with just some kind of pleasant mood or setting in the background, the listener gets to hear, when all of these various ideas come together, that the song is truly greater than the sum of its parts. Secondly, instead of making the music sound "dry," it actually becomes more expressive, since the pianist must now seriously consider the phrase and the interrelationships of the parts, provided that the pianist has, in fact, learned the art of genuine legato playing. The result is that the performers will possibly discover new insights into the construction of a song, and, therefore, more fully realize the composer's intentions

You have to be a member to access this content.

Please login and subscribe to a plan if you have not done so.