Introduction
This essay will focus on discussing the effectiveness of the choice of tempo and tempo rubato employed by the performer. Tempo is a considerable aspect of all music; although appropriate tempo choices do not guarantee a successful performance, misinterpreting the composer’s tempo markings will markings will certainly distort the character and intention of the piece. Any piece played in strict time without any form of freedom in tempo is likely to sound rigid and uninteresting. To aid in an engaging performance, flexibility of tempo within the music, or tempo rubato (literally ‘robbed time’) is applied frequently regardless of any indication by the composer, and used in fairly subtle ways to enrich the musical phrase (Costa, 2012: 251). For this research I have selected Frédéric Chopin’s Barcarolle Chopin’s Barcarolle (1845-46) recorded Norwegian pianist Håkon Austbø Aus tbø.. Håkon Austbø is a professor at the Amsterdam Conservatoire, and specialises in the works of Messiaen and Skryabin. He has collaborated with artistes of other disciplines with uncommon repertoire and his discography includes piano music of Brahms, Debussy, Grieg and J anácek. Chopin’s Barcarolle Chopin’s Barcarolle
A Barcarolle imitates songs sung by Venetian gondoliers, and features a marked lilting rhythm depicting the gentle rocking of a boat. Composed in 1845, Chopin’s Barcarolle was one of his last major published work, which features a 12 time signature instead of the 8 6
customary 8 .The main edition used for reference in this study is the Polish National National Edition, with Jan Ekier as the editor. An average performance lasts about eight and a half minutes.
Tempo
Tempo markings are used to differentiate sections within a movement, or connect works on a deeper level; the markings in the Barcarolle carry a similar importance (Hood, 2014: 196). Tempo, dynamic markings and other performance directions have to be read within the context of some understanding of common practices and influences during the period of composition. An appropriate tempo is one that falls within plausible parameters rather than a defined absolute tempo, and the performance should stick close to the chosen tempo until another is indicated (Davies, 2001: 59). In Koch’s Musikalisches Lexicon (1802: 62) , it is stated that tempo terms can be used to indicate the speed, or merely the style of the performance, or both. Chopin viewed his performance markings as examples on the direction of the performance, rather than fixed prescriptions (Butt, 2002: 112), and often made changes to his tempo marking. In some works, the metronomic markings are said to be more credible when performing on a Pleyel due to the lighter keys as compared to the modern piano. In his earlier works we can find exact values for the tempo along with Italian tempo directions, and by 1836 Chopin has eliminated his tempo markings altogether (Jackson, 2005: 88). From these earlier works it is possible to deduce and establish a range of limits for the tempo of his later works but there also limitations when attempting to do so. Chopin’s metronome markings for Allegro ranges from to
= 192 (Mazurka Op. 24 No. 2), and Allegretto from
___ = 60 (or
= 92 (Etude Op. 10 No. 9)
= 76 (Etude Op. 10 No. 11) to
= 180: Menuetto from Sonata No. 1 in c minor, Op. 4). These extreme ranges
complicates the method of deducing the tempo for later works. Assuming the tempo markings were authorised by Chopin, we can conclude that Chopin saw radical differences between the
various genres (Higgins, 1973: 108) and that tempo directions could be seen more as a characterisation or performance direction of the piece. Chopin’s Rubato
In the 18th century, rubato may be notated by grouping and rhythmically diminishing notes, but it was to imply the idea of an accelerando instead of a strict rhythmic diminution (Figure 1). By the 19th century, composers began notating rubato, and the practice of rubato was applied to both the melody and accompaniment (Gunn, 2015: 137).
Figure 1: Notated Rhythmic Acceleration
One of the most distinctive musical qualities of Chopin was his two types of rubato. The first kind of rubato was found to be more common, which is described as the ‘agogic modification of pace relative to the basic tempo’. The duration of the rubato varies, but the variation takes place simultaneously in both melody and accompaniment (Kallberg, 2004: 243). However, this form of rubato should not skew the time taken for the performance, as rubato should not only comprise of stealing time (accelerando), but giving back (ritardando) as well. The use of accelerando and ritardando heightens drama noticeably. Accelerando can be used when ideas are repeated more intensely, or unexpectedly to arouse a more passionate effect. Ritardando is more common in passages that suggests longing and melancholy, and is often used regardless of indication by the composer, with notated ornamentation, approaching fermatas, transitions or when ending a passage or phrase (Gunn, 2015, 138). Where one part of a phrase is quickened, another should slacken in proportion, in order not to disturb the general pulse; the duration of the phrase played with rubato should thus remain the same if
played in strict time throughout (Taylor, 1897: 73). This ‘older’ concept of rubato still conforms to Davies’ views, as expressed earlier. The second, described by his pupil Wilhelm von Lenz is the rhythmic variation occurring in the melodic lines while the accompaniment remains rhythmically constant (Eigeldinger, 1986: 49), is similar to Mozart’s style of rubato as accounted in his letter from 1777. In a study of music performances of Bach and Scarlatti, both ‘historically-informed’ and ‘mainstream’ performances were alike in the way that the performers employed tempo alterations cadences and during significant thematic, harmonic, rhythmic or structural moments. The only distinction was the historically-informed performers kept a solid rhythm in the left hand bass, which matches the described style of Chopin’s rubato (Ornoy 2006, 237). Coined as metrical rubato by Costa, this old bel canto type of rubato is comparable to an opera singer retarding and accelerating at almost every moment while the orchestral accompaniment proceeds steadily (Christiani, 1885: 301). This latter form of rubato inevitably leads to the discussion of Dislocation, which is the art of playing one hand after the other. There is much evidence of the practice of dislocation in early recordings by Leschetitzky, Saint-Saëns and Paderewski (Costa, 2012: 47). The melody-notes may be struck an instant after the bass, which gives it a softer effect; dislocation should be done only at beginning of phrases, important notes and strong beats. The delay should be minimal to make it ‘hardly noticeable for the uninitiated’ (Brée , 1902: 73). When properly executed, dislocation serves to emphasise or de-emphasise the affected melody note(s), not necessarily to bend the rhythm of the melody. This practice is now met with criticism, as it was possibly the ‘apogee of expressiveness’ in the 19 th century that has been discontinued with time (Costa, 2012: 45, 100).
Having described the various aspects of tempo and tempo rubato we shall now look in detail the recording of Chopin’s Barcarolle by Austbø, in an attempt to determine to what extent his recording was influenced by historical practices. Not all instances will be highlighted due to the word limit and possibility of repetition. All tempo markings are measured by tapping on a digital metronome, and verified with the timings on the recordings. The tempos are also an approximate value due to the amount of rubato used in the music.
Austbø’s Rendition of the Barcarolle
The performance started with a commanding forte, where the corresponding dynamics was found in the first English edition based on the revised autograph edition which is currently lost. Austbø fluctuates the tempo significantly at the opening phrase, with a stretched first bar at
= 40, picking up the tempo to
with a ritardando to almost
= 60 at the second bar and ending it
= 18. There is also a slight tenuto on the first quaver beat of the
second bar, where emphasis is given due to an elongated rhythmical value. Austbø then compensates this alteration by shortening the value of the preceding quaver. The dotted minim rest was also marginally shortened before the accompaniment enters in bar 4.
In the following table I have tabulated the tempi directions by tapping the tempo on a digital metronome for the first reading, and divided the time taken to play the various passages by the number of beats for the second reading. All metronome readings are in dotted crotchets per minute. Bar number
Tempo Indication
Austbø
Austbø average
1
Allegretto
60
52 (56)
35
poco più mosso
60
57
62
poco più mosso
72
63 (65)
72
meno mosso
50
51
84
tempo primo
62
55
93
più mosso
70
59 (65)
103
tempo primo
50
52
Table 1: Tabulated tempo in Austbø’s performance.
The bracketed readings were taken by omitting the last bar(s) of the section due to the amount of ritardando applied, by doing so we see the average tempi would then be slightly closer to the tapped tempi. There is a general contour for the tempo relationships marked by Chopin, and Austbø choice of tempo follows Chopin’s tempo directions to a certain extent. The exceptions happen at the first poco più mosso section beginning at bar 35 where the tempo is kept from the previous section without any significant change, and the tempo primo at the coda where he took a significantly slower tempo as compared to the tempo in the first section.
Christiani categorised tempo terms into five fundamental degrees: very slow, slow, moderate, fast and very fast; the lower threshold for fast being Allegro, and Andante for moderate (1885, 261). With Allegretto being the diminutive of Allegro, ‘moderately fast’ would be an approximate description. Conventional metronomic markings will not work: at Allegretto it would be too fast when counting in dotted crotchets and to slow when counting in quavers. As discussed earlier, it would be more appropriate to view it as a performance direction, instead of an absolute value. In late eighteenth century, Allegretto was understood to be a somewhat fast tempo, without showing any sign of hurry (Fallows 2001, 382). Austbø captured the depiction of a boat song relatively well, as with other convincing recordings of performances, with tempos ranging from Arrau (Chopin, 2011) at 2006) albeit on the faster end at
= 55 to Cortot (Chopin,
= 70.
I have measured the tempo for the first section in the following ways: Tapping the pulse on a digital metronome per dotted crotchet, per dotted minim, and mapping the time taken to play 2-bar and 4-bar melodies. The methods produced varied results. The first method showed an unsteady pulse varying from
= 40 to 65. This is result is somewhat
supported by the third method, where the time taken to play each sub-phrase differs by 3 to 15%. The second showed a more steady pulse of
= 24 to 30, which in turn accords to the
fourth method, where the difference in time varies from 2 to 10%. Austbø is constantly applying rubato possibly in a bigger context, which matches the types of rubato described earlier, albeit with his own modifications. The rubato is applied to the elaborated accompaniment notes within the dotted minim, so the listener is still grounded by the bass note struck on every 1st and 3rd beat of the bar. The underlying accompaniment thus appears to be kept at a steady pulse, preserving the character of the Barcarolle.
The occurrences of rubato is also evident when trills are present in the melody. The following examples highlights the lengthened notes.
Figure 2: Lengthened notes due to presence of trills
Metrical rubato is slightly apparent in Austbø’s interpretation, as he kept fairly true to the rhythmic values that were notated, while altering marginally the rhythmic values at certain moments. For example, this build up before the extended authentic cadence (figure 3) and other instances of musical tension (figure 4).
Figure 3: Bars 28 – 29, build up before authentic cadence.
The semiquaver beats (boxed) in the melody in bars 28 and 29 are slightly lengthened, sounding almost like a triplet. Also, the semiquaver beats (circled) in the accompaniment are played faster than notated, which gives an impression that the right hand chord is held longer than usual. The accelerando in the semiquavers is then compensated by a ritardando in the preceding beat.
Figure 4: Bar 26, rubato in trill.
In bar 26, a ritardando is gradually applied during the execution of the trill, in an attempt to increase the build-up in tension before the leap to the appoggiatura in bar 27. This conforms to the earlier mentioned practice of applying rubato in notated ornamentation. The cautious use of metrical rubato must be observed from bar 51 due to the following reasons noted from the compositional style: i.
The ostinato rhythm in the left hand has been in place from bar 39.
ii.
The chordal texture in the right hand augments the above mentioned rhythm.
iii.
The first quaver beat(s) in the left hand is doubled by an octave.
The above reasons are contributing to the impression of an accent on every first beat, for which in Austbø’s case, is further exaggerated with his augmentation of every first quaver in each beat. When metrical rubato is applied too often it disrupts the line, making the music sound fragmented. There are a few instances where Austbø stretches the first quaver of every beat, but it is most noticeable at these sections.
Figure 5: Bars 51 – 53, emphasis on first quaver of every beat.
Figure 6: Bars 30 – 31 and 90 – 91, emphasis on first quaver of every beat.
In this case although the rubato is applied appropriately at musically significant moments (build up before cadences) with the accompaniment pulse being steady, there is significant distortion in the rhythmical values, especially of the left hand. Metrical rubato as discussed earlier is usually applied more often to the melody notes, and in Austbø’s performance is happening more in the accompaniment than the melody. There are two occurrences of fermata in Chopin’s Barcarolle, with almost similar chordal progressions leading to a semiquaver rest which is held by a fermata. Both figurations lead to different sections, thus the different dynamic markings.
Figure 7: Bars 32 and 92, chordal progression leading to fermata.
When rubato (accelerando and ritardando) is used sparingly at the suitable times, with correct spacing within the notes, it augments the tension already created harmonically by the composer. In this section the fermata on the semiquaver rest, which is an expression of
surprise or astonishment, further emphasises the tension (Gunn, 2015: 138). In bar 32 Austbø applied slight accelerando to the first half of the bar, and ritardando in the second half; in bar 92 the effect is amplified by extending the accelerando until the more substantial ritardando in the last quaver and semiquaver beat. Austbø clearly demonstrated understanding of the structure and transition through skilful use of the rubato. Dislocation is less conspicuous when done appropriately. Austbø’s displaced melody notes are subtle enough to not be noticeable upon the first hearing, yet sufficient to produce the effect of softening the melody note. The first example of dislocation happens in the first phrase, on the highest notes of the phrase (circled in the diagram: D#, F# and G#, B), which observes the dynamic markings of the phrase. The subsequent images highlights other examples where dislocation occurs, usually heard at the end of a phrase.
Figure 8: bars 6 – 10, dislocated melody notes
Figure 9: bars 15 – 16, dislocated melody notes
The most persistent use of dislocation is in the ending più mosso section, applied frequently but cautiously.
Figure 10: bars 94 – 97, dislocated melody notes
Austbø’s use of appropriate dynamics allowed him to soften and de-emphasise the melody notes in the first two examples (figures 8 and 9), and emphasise the melody notes in the more climatic section leading to the coda. Although Leschetizky suggests dislocations should be imperceptible and weaker melody notes should coincide with the beat, his recordings suggests that he employed dislocation at several other moments with obvious delays (Brée 1902, 73; Costa 2012, 73). From his recording of his Ballade Venitienne (Barcarola) (Chopin, 1992), the first beat is always displaced, and the descending figuration is also displaced, and to a certain extent arpeggiated. In his recording of Chopin’s Nocturne (ibid.), the dislocation also happens far more often and perceptible, with reference to his teachings.
Figure 11: Ballade Venitienne (Barcarola) bars 1 – 6, dislocated melody notes
Figure 12: Nocturne in D-flat Major, Op. 27 No. 2, dislocated melody notes
According to Chopin’s pupil Carl Mikuli, Chopin firmly opposed to the practice of appeggiation of chords; unlike Brahms who arpeggiated most chords when he played according to witnesses (Rosen 1995, 413). These practices appear to be popular with pianists born in the latter half of the 19th century including Alfred Cortot, Myra Hess, Sergei Rachmaninoff and Ernö Dohnanyi (Costa 2012, 47). In the final tempo primo section, Austbø’s arpeggiation is apparent in the octave bass notes, along the different elements of rubato discussed earlier. Conclusion
There are many other aspects of historically informed practice that were not discussed due to the focus on tempo and tempo rubato. These include, and is not limited to: rhetorical rests or musical punctuations (another subsidiary of tempo rubato), positioning of melodic decorations before or on the beat, doubling of bass notes and interpretation of the extrametric groupings (septuplets and quintuplets in bars 78-80). Along with the evolution of the piano, these performances practices have also changed over time, and what we practice today may be radically different with the norms in the common practice period. Austbø’s performance of the Barcarolle, which appeared to be unpretentious on the first hearing, revealed a number of historically-informed elements upon analysis. There is always a conscious effort on the application of rubato, where he compensates accelerando with an equal part of ritardando. Metrical rubato is also applied to a certain extent, however differing to 19th century practices at some points, where it occurs more in the left hand accompaniment than the stated right hand melody. Austbø’s application of dislocation were rather inconspicuous but adequate to serve the purpose of ‘softening’ the melody notes. Interestingly, his method of displacing notes are more appropriate to Leschetizky’s description of dislocation, than Leschetizky’s own practices as heard from his recordings.
Although there might be external influences affecting the performance, for example the acoustics of the recording location (recording was made in Østsiden Church) and the state of the instrument, from the observations and analyses made through the essay we can conclude that to a large extent, Austbø has taken into account the performance practices current at the time of composition.