Baroque harmonic language and chord playing on the marimba: one piece, many options

Read as PDF

DOI: 10.32063/0609

Tomoyo Ueda

Percussionist Tomoyo Ueda’s special interest lies in the performance of early music on the marimba. She has given concerts, masterclasses and lectures on the subject in Europe, Singapore and Australia and her paper Marimba Plays Early Music was published by the Imperial College Press, London.

by Tomoyo Ueda

Music & Practice, Volume 6

Resistance and Affordance

Works from the Baroque period form an important part of the marimba repertoire. Performing works such as the Cello Suites and the Sonatas and Partitas for Solo Violin by Johann Sebastian Bach gives marimbists the opportunity to delve into repertoire from outside the 20th and 21st Centuries, when original works for this instrument began to be written. This practice of appropriating earlier repertoire from string instruments has become so common that works from the Baroque period frequently appear on audition lists for music schools and orchestras, as well as on concert programmes by students and professional soloists. However, performance of Baroque music presents marimbists with some difficulties.

In the case of original works for the marimba, and in keeping with the notational conventions of these works’ relatively recent dates of composition, various detailed instructions, such as precise tempi (metronome markings), dynamics, phrasing and articulation, are usually written in the score, except when the composers withhold such comprehensive information, specifically instructing performers to improvise or interpret what is notated on the page. Such an instruction may be supplemented by explanatory notes or may make use of graphic notation, upon which the performer is invited to create their own musical passages within the piece. In this way, even if performers do not know the biography of the composer or the context in which the work was composed, they can generally utilise simply the information and instructions written in the score to produce a reasonable interpretation. By contrast, in Baroque music such detail is often absent from the page. Many scores offer the performer little information other than the notes and, perhaps, a general indication (Allegro, Presto, Adagio, etc.). It is up to the performer to ‘flesh out’ the score by adding elements such as dynamics and ornamentation. By doing so, a performer may produce an interpretation that is unique to them. For this reason, many marimbists prefer to use material in the form in which it was conceived, and therefore still scored for the instruments for which these works were originally written, instead of employing a transcription of the work in question for the marimba, in which various performative instructions have been inserted by modern editors.

Some marimbists use Urtext editions, such as those by Bärenreiter, and take the lack of instructions on the page as an open invitation to interpret the work ‘freely’. They are often unaware that, although each interpretation should indeed be different and individual, decisions made by performers should be based on a knowledge of the conventions, rules and guidelines of performance practice as these apply to the Baroque era and style. Some reject this idea and make a conscious decision to create their own interpretations that are based on their intuition or knowledge, which stem from their experience of performing original repertoire for the marimba. As a result their interpretations often do not reflect the Baroque style. Others are simply at a loss as to what to do because they do not have the knowledge or experience to comprehend what is not on the page – that is to say, the implicit parameters of appropriate performance practice that would have been instinctive to performers active at the time of the work’s composition and which therefore did not need to be spelled out to them. In an attempt to compensate for their unfamiliarity with historical performance practice, some marimbists consult Baroque specialists for advice about the works that they are studying and they also listen to recordings of them made by such specialists.

While I agree with the approach of listening to recordings and receiving instructions from Baroque specialists, I often feel that marimbists fail to investigate the Baroque style in general before studying a specific work for performance. In my case, when performing works from the 18th-century, I approach my studies from several angles.  During my percussion studies, I also took Historical Performance Practice courses. Apart from receiving instructions from Baroque specialists and listening to recordings of the works that I am studying, I listen to recordings and live concerts of not only the works that I am studying but also of other related works so as to immerse myself in the relevant sound world. Texts such as treatises by C.P.E Bach or J.J. Quantz provide me with vital information about the conventions and practices of the period. And performing not only solo and chamber works but also orchestral repertoire gives me the opportunities to put the information that I have gathered into practice.

This process of generalised familiarisation with Baroque performance practice prior to addressing an individual work has been vital for me when studying Baroque music on the marimba. It has also, I believe, enriched my teaching. For example, when I gave a masterclass in Germany about the performance of early music on the marimba I was asked by students how one should insert and perform Baroque ornaments in the Cello Suites by Johann Sebastian Bach. Although it would have been possible to go straight to illustrating different possibilities of ornamenting particular passages in the Suites, I felt that there were other more general aspects of performance practice that needed to be considered before deciding how to ornament specific passages. Among these, I would include the characteristics of different Baroque dance movements and their rhythmic structure.

For example, when studying a minuet it is important to know that it is a dance with a moderate tempo in ¾. The rhythmic structure consists of a strong 1st beat followed by a weak 2nd beat and the 3rd beat as the upbeat to the next bar. This information should be taken into account when making performance decisions such as ornamentation. It suggests that it may be inadvisable to ornament a note or figure that is played on the 2nd beat as to do so may bring unnecessary attention to a weak beat. Rather, it may be more appropriate to insert a trill on a note on the 3rd beat so the passage could lead smoothly into the next bar. Meanwhile, the sarabande is another dance in triple metre but slower than the minuet and, this time, with an emphasis on the second beat. Ornamenting this beat would therefore make musical sense. As seen later in the paper, this question of the dance origins of many Baroque compositions should kept in mind when considering how to play chords in these and other movements from suites and partitas.

But the importance of contextual knowledge goes wider than this. As well as being acquainted with the characteristics and aspects of historical performance practice before studying a specific piece, when performing Baroque music on the marimba I believe that one must also have some understanding of Baroque thinking and ideals. For example, a quintessential element of Baroque music is that it expresses affects or passions. These affects may be understood as ‘organized emotion[s]’[1] or as ‘generic states’[2]  that codify and correlate human feelings with musical features in a way with which both musicians and audiences at the time would have identified. Musicians were required to learn and master the rules and guidelines of performance as to the use of harmonic language, rhythm and ornaments and then combine them with their ‘good taste’ to produce a unique interpretation which nevertheless appropriately expressed these affects. They combined their knowledge with their personal intuition to achieve the ultimate goal: to ‘move and please’ through making the audience feel the affects expressed by the music and causing them to be moved by them.[3]

In my opinion, one must consider how one can incorporate this concept of the ‘doctrine of affections’ into the performance on the marimba of works that would have been written according to its principles. By combining my knowledge of historical performance practice, gained through study and research, with my technique on the marimba, I aim to produce a stylistic interpretation that achieves the goals pursued by Baroque musicians while, at the same time, being uniquely expressed through the characteristics of my instrument. I will illustrate my argument using examples from the Suite for Violoncello Solo No. 2 in D minor, BWV 1008 by Johann Sebastian Bach. The purpose of this is not to dictate how others should play early music on the marimba. Instead, I shall attempt to show different possibilities of expressing different affects using particular characteristics of this instrument, thereby encouraging each marimbist to create his or her own individual performance.

One important aspect of performance practice is the composer’s use of harmony. It is a driving force in 18th century music, enabling it to express different affects. By using a particular key the composer introduces the main affect of the piece. Then, by alternating between dissonance and consonance and also through the use of harmonies that are contrary to the predominant affect, the composer may be able to lead the performer and audience away from and back to it in ways that give specific colours to the piece and enhance its overall impact.[4]

Armed with this understanding, the performer might first examine the bass part, harmonic progressions, dissonances and resolutions, considering what they mean and what functions they have within the piece or passage. Then he or she should ask themselves how one could present them on the marimba so that the audience might recognise the affects that are being expressed. For example the Gigue from the Cello Suite No. 2 begins with a strong statement in D minor (bars 1-4), modulating to F major (bars 5-8). The turbulent, aggressive opening created by the tonality and leaping intervals (bars 1-2) contrasts with the milder, gentler section that leads us to the relative major (Fig. 1). To demonstrate the contrasting tonalities one could, for example, use different dynamics: bars 1-4 could be played loudly and strongly with full strokes and the modulation to F major could be played more quietly and with softer articulation (Audio Example 1).

Fig. 1 J.S. Bach, Suite for Violoncello Solo No. 2 in D minor, BWV 1008, Gigue, bars 1-8

In addition to contrasting dynamics, bars 1-4 could be played ‘as written’, that is to say rhythmically ‘straight’, while rhythmic inflexion and appogiatura could be used in bars 5-8.

The variation in rhythm would provide further contrast between different affects (Audio Example 2).

In order to make performance decisions such as these, a basic knowledge of Baroque harmony is vital, as is an appreciation of the practice of playing a dissonance more strongly than its resolution and of playing the bass note of a chord on the beat (as opposed to before the beat and the melody on the beat, as frequently done in the music of later periods). For example, in bars 9 and 10 of the Sarabande from the Cello Suite No. 2, a consonance on the first beat is followed by a dissonance on the second beat (Fig. 2). The placement of a dissonance on the second beat corresponds to the rhythmic structure of the Sarabande, which, as discussed earlier, has an emphasised second beat. The tension caused by the dissonance is resolved by the consonance on the following downbeat. For me, the practice of playing the bass note on the beat and thus starting the chord on the beat displays a clear rhythmic structure, assisting the performer and the audience to understand the content of the music.

Fig. 2 J.S. Bach, Suite for Violoncello Solo No. 2 in D minor, BWV 1008, Sarabande, bars 9-10

A crucial component of performing earlier music on the marimba is the exploitation of the instrument’s characteristics. A prominent feature of the marimba is its large dynamic range. It is also not as difficult to control dynamics as it is for many other instruments, as intonation is not affected by the change in dynamics, and one could easily shift very quickly from, for example, ppp to fff  by changing the speed and height from which the mallet falls to strike the bar. Another aspect of the marimba that differs greatly from bowed string instruments is that, on the marimba, one can play up to four notes simultaneously (when playing with four mallets) and all notes can be played clearly and cleanly. It is also easy to emphasise one or more notes in a chord when multiple notes are played at the same time. However one cannot sustain notes on the marimba except when the note is played as a roll. On an instrument such as the violin or cello, it is easy to sustain notes using the bow, but harder – especially on the modern instrument with its more rounded bridge and tauter bow – to play more than two notes simultaneously.

Because the marimba and bowed string instruments differ greatly in their tone qualities and characteristics, instead of blindly copying performances by musicians for whose instruments these works were composed, I prefer to experiment on the marimba and consider how I can apply the unique qualities of this modern instrument to highlight ‘the true content and affect of a composition’.[5]  Such a process can be applied, for example, to the beginning of the same Sarabande discussed above (Fig. 3). The movement starts with a stark tonic-dominant-tonic harmonic progression in root positions (bars 1-2). The progression once again corresponds to the rhythmic structure of the sarabande, as the stable tonic is placed on the first beats and the strong dominant, which pulls the listener back towards the tonic, on the second beat. Bars 5-6 follow the same harmonic progression and rhythmic structure as bars 1-2; they also feature the same melody in the top voice. However, the phrase in bar 5 begins in the sweeter, less stable-sounding first inversion of D minor with F in the bass.

Fig. 3 J.S. Bach, Suite for Violoncello Solo No. 2 in D minor, BWV 1008, Sarabande, bars 1-6

Although bars 1-2 and 5-6 are almost identical, I find that bars 5-6 could sound less intense and more introverted than bars 1-2 due to the D minor chord in the first inversion in bar 5. For me it is important to bring forth the different characteristics and nuances of these phrases by playing chords in the passage in different ways. For example one could play all notes in the chords simultaneously or as arpeggios, and play the arpeggios faster or slower. In addition one could split the chord, for example playing the bass first and all the other notes in the chord later.

At the beginning of the Sarabande, the notes in each chord in bars 1-2 could be played simultaneously, making a strong, definite statement suitable for this serious, noble movement. The chords in bars 5-6 could then be played as arpeggios. Spreading out the notes in the D minor chord in bar 6 enables the notes to fill out the space of 1½ beats. Playing bars 5-6 softer and with less articulation than bars 1-2 generates an echo effect, which, to my mind, is in keeping with the use of the first inversion chord in the written score (Audio Example 3).

It is also possible to play all the chords as arpeggios. This would enable all notes in the chords to be clearly audible. However, the contrast between chords and single line melody (bars 3-4) would then become less pronounced and the passage as a whole would become rounder (Audio Example 4).

Alternatively, the passage could be made to sound more angular and aggressive by playing all notes in the chords simultaneously, with similar articulation and in strict tempo. Again bars 5-6 might be played more softly than bars 1-2 to establish their different character (Audio Example 5).

Returning to bars 9-10 of the Sarabande (shown again below as Fig. 4) the marimbist could consider which notes of the dissonant chords should be emphasised. For example, in bar 9 on the second beat one could play the notes of the chord simultaneously while bringing out the B-flat and A to highlight the dissonant interval of major 7th, which resolves harmonically in the next bar on the first beat. Similarly, in bar 10 on the 2nd beat one could play the G and F simultaneously, again to accentuate the dissonance. The placement of the dissonance on the second beat of the bar corresponds to the rhythmic structure of a sarabande, emphasising the second beat. This is sweetened by the B-flat which could be played separately and later than the other notes in the chord (Audio Example 6).

Fig. 4 J.S. Bach, Suite for Violoncello Solo No. 2 in D minor, BWV 1008, Sarabande, bars 9-10

When chords are played as arpeggios, every pitch becomes audible. By considering where to place each note and which dynamics to use, it is possible to bring attention to certain notes in the chords and to highlight their function. For example, one can bring out the harsh major 7th interval on the second beat of bar 9, created by the B-flat and A by playing the chord as a fast arpeggio and emphasising the A in the melody. In contrast, the chord on the second beat of bar 10 could be arpeggiated more slowly, played less intensively and with the B-flat placed late to bring attention to its sweetness (Audio Example 7).

Alternation between dissonance and consonance occurs again in bars 25-26 of the Sarabande. However, tension also mounts over two bars as the bass line rises chromatically (Fig. 5). The performer could choose to focus on the shifts between tension and resolution caused by the diminished 5ths on the second beats of the bars, each of which is resolved by the consonance on the first beat of the following bar. Alternatively, the marimbist could highlight the bass line, which creates a single arc of increasing suspense across both bars.

Fig. 5 J.S. Bach, Suite for Violoncello Solo No. 2 in D minor, BWV 1008, Sarabande, bars 25-26


As I have tried to show, when studying Baroque music on the marimba it is important to utilise not only the performer’s technique, instinct and ‘good taste’ but also every piece of available knowledge about historical performance practice and the style of the Baroque period. Once the performer understands this, he or she can experiment on the instrument and even two bars of music may become full of fascinating interpretative questions and possible answers. Of course, all performances should be broadly in accordance with the conventions of historical performance practice and style. However, as stipulated by C.P.E. Bach,[6] every performance should also be unique to the performer (to which I might add it should be sensitive to the unique properties of the instrument on which it is performed). For me this ‘balancing act’ that the performer undertakes is the greatest attraction of performing 18th-century music on the marimba.

Of course, the marimba as a western classical instrument did not exist in the Baroque period and its properties are quite different from those of the instruments for which works such as the Cello Suites of Bach were composed. No performance on the marimba of a Baroque piece originally written for a bowed string instrument is going to sound literally as the composer would have envisaged it. In my view, the marimbist’s task is therefore not so much one of imitating as of re-imagining. In a similar way to a pianist performing on a modern grand piano compositions originally conceived for harpsichord or organ, the challenge is to find within the music qualities that, while already present in latent form, are only brought out fully when combined with the specific properties of an instrument unknown to the composer. The justification for using an anachronistic instrument is then that it may add another dimension to music which, certainly in the case of Bach, is sufficiently rich and profound to carry a multitude of interpretations. The piano cannot take on the harpsichord or organ on their terms; however, it does have prodigious capabilities which they both lack. By understanding the broader principles of Baroque performance practice, the pianist – or, in my case, the marimbist – is more likely to find ways of exploiting their instrument’s unique potential in a manner that honours not just the historical context within which the music was originally composed but also the motivating spirit behind it.

In this article, I have suggested ways in which the marimbist might use the instrument’s unique strengths to bring out in novel ways qualities already present in Baroque music. By doing so, and despite the fact that they are performing on an instrument which only became properly recognised within the classical music world nearly 200 years after this music was composed, they can still aspire to move the audience, just as Baroque musicians did, through the affects embedded in the music. This is why, for me as a marimbist, it is important to research about historical performance practice when studying a work from this period. Once the framework is established within which the marimbist is to work, he or she may realise that, although there are rules and guidelines that should be followed, there are still myriad possibilities in terms of how precisely to perform this music. I believe that, rather than reducing these possibilities through the imposition of rules and restrictions, knowledge of historical performance practice can open them up, multiplying the options for the performer and enriching the experience for the audience.


Bach, Carl Philipp Emanuel, W. J. Mitchell, ed. & trans., Essay on the True Art of Playing Keyboard Instruments (London: Ernst Eulenburg Ltd., 1980)

——-   W. Horn, ed., Versuch über die wahre Art das Clavier zu spielen (Kassel: Bärenreiter, 1994)

Haynes, Bruce & Geoffrey Burgess, The Pathetick Musician: Moving and Audience in the Age of Eloquence (New York: Oxford University Press, 2016)

Kuijken, Barthold, The Notation is Not the Music: Reflections on Early Music Practice and Performers (Bloomington, Indiana: Indiana University Press, 2013)

Quantz, Johann Joachim, Versuch einer Anweisung die Flöte traversière zu spielen (Kassel: Bärenreiter, 2000)

——-   E. R. Reilly, ed. & trans., On Playing the Flute: The Classic of Baroque Music Instruction (London: Faber and Faber, Ltd., 2001)

Tarling, Judy, Baroque String Playing for Ingenious Learners, (Hertfordshire: Corda Music Publications, 2000)


[1]    Barthold Kuijken, The Notation is Not the Music: Reflections on Early Music Practice and Performers (Bloomington, Indiana: Indiana University Press, 2013), p. 107

[2]     Bruce Haynes and Geoffrey Burgess, The Pathetick Musician: Moving and Audience in the Age of Eloquence (New York: Oxford University Press, 2016), p. 8

[3]    Johann Joachim Quantz, E. R. Reilly, ed. & trans., On Playing the Flute: The Classic of Baroque Music Instruction (London: Faber and Faber, Ltd., 2001), p. 23

[4]     See Haynes and Burgess, p. 92

[5]    C. P. E. Bach, W. J. Mitchell, ed. & trans., Essay on the True Art of Playing Keyboard Instruments (London: Ernst Eulenburg Ltd., 1980), p. 148

[6]    See C. P. E. Bach, W. Horn, ed., Versuch über die wahre Art das Clavier zu spielen (Kassel: Bärenreiter, 1994), p. 117