Exercise in Styles:
On Developing Multiple Versions of Debussy’s Syrinx

The indications related to the tonal parameters of vibrato, dynamics and articulation provided by the score are either relative, as in the case of dynamics, limited by the system of notation, as in the case of articulation, or completely absent and left for the discretion of the performer, as in the case of vibrato and portamento.


Looking into vibrato, the Syrinx score doesn’t provide any instructions and the common practice is to use continuous vibrato type that was originated from, and is strongly associated with the French School of flute playing. This vibrato type became dominant globally from around 1950 onwards, and the other vibrato types that were used during the first part of the century practically disappeared. Taking an historical perspective, the use of the French School approach to vibrato in Debussy’s Syrinx seems an obvious choice. An interesting question, however, is how Syrinx would sound like if played using other approaches for vibrato, or with no vibrato at all?

The departure point for the first version was a non-vibrato approach. Today this approach is widely used and discussed in the context of historically-informed performance practice of 18th and early 19th-centuries repertoire. In the context of early flute recordings, it was one of the neglected approaches to vibrato, which many of the German flutists applied to any repertoire. From that perspective, using a non-vibrato approach in the context of Syrinx can’t be justified. The result, however, sheds new light on the expressive range that becomes available by approaching Syrinx in a slightly non-conventional way. The outcome is emotionally restraint, and may fit the ‘Très modéré’ instruction of the score. More generally, it may be found suitable for the expressive ideal of French music of the period as opposed to the overt emotionalism associated with German music of the time.

Going to the opposite extreme, in the following version I explored using vibrato type inspired by the one found in historical recordings that were made mainly in England during the first part of the 20th century. This type of vibrato is characterised by a faster vibrato speed and smaller width. Unlike the modern vibrato, in which the undulations take place in both the loudness and pitch, and the production involved using both throat and abdominal muscles, in this type of vibrato the undulations take place mainly in the loudness, and it is produced by narrowing and opening the throat. This vibrato and the sonority associated with it were rejected by the French School players and their followers, and taking an historically informed perspective, this approach cannot be justified. The result, however, sheds a new light on what is afforded by Debussy’s score.

As I was exploring the use of this type of vibrato, the emotion that came to my mind is of panic, which results to a large extent from the throat production of the vibrato, and which reminded me the vocal tremble that appears in extreme emotional states. Panic is of much relevance in the context of the mythical story of Syrinx and Pan, and the association with extreme fear served as a departure point for constructing the performance narrative in this version.


The focal point of the next version is dynamics, which in most scores take the form of relative indications regarding the level of loudness. It may also include verbal metaphors referring to character and overall intensity, rather than to the volume of sound. The range of dynamics changes indicated, and the frequency in which they appear, are an important information for performers for extracting meaning from the score’s instructions. The relative nature of the dynamics instructions makes it negotiable, depending on the size of the space used, its acoustics, the instrument and the physical condition of the performer on the day of the performance.

The subjective nature of the terminology used to indicate dynamics also leaves performers with margin for personal reading. In the case of Jobert’s edition, the dynamic range indicated is between p and mf, which in some of the cases are followed by crescendo and diminuendo, and in which the ending point in terms of loudness is not indicated. The very end of the piece, the indication is for an overall effect of ‘dying away’.

The score’s indicates a mf opening, which may serve as a reference for the level of loudness that follows. But how loud is mf? In this version I explored starting with a dynamic level that is quieter than it is usually done. As expected, the other performance parameters are also effected: the attacks are softer, the vibrato is more shallow and discrete, and tone colour needed to be adjusted by using harmonic fingerings in several places. Maintaining the quiet dynamics made this version quite technically demanding, and as a result, the performance is intimate and dreamy but with introvert tension.


The tonal parameter of articulation is a complex one in the sense that it is an aspect of timing that is perceived as a tonal parameter. Articulation is also closely related to dynamics and tone colour, is very difficult to be measured and analysed using software for sound analysis, and received very little attention in the study of musical performance. The articulation continuum spans from full legato through different degrees of separation between the notes, and relates to both the way the note begins as well as the way it ends. The wide range of possibilities that is available in performance cannot be covered by the symbols used in notation, and the decision how to shape articulation remains to a large extent within performers’ domain.

In Syrinx score, most of the notes are tied together in small groups. Probably following this wide use of legato in the score, Syrinx is usually performed with soft attacks and little separation between the end of the notes and the following ones, resulting in long and continues lines. In the next two versions, the focal point is on the quality of the attacks and on longer separation between the end of the legato notes and the note that follows. This results in a dancing like character: light in the first version, and more aggressive in the second one.