An arpeggiation that speeds up and slows down once or several times, often evokes the association of a visual gesture. Consequently, this section reflects on how the internal tempo structure of an arpeggiation can be translated into a gesture-like visualization.

In 2008, a team of authors researching in Austria proposed using phase-plane visualizations to emphasize the dynamic and gestural quality in expressive timing (Grachten 2008) and slightly extended it in 2009 (Grachten 2009), highlighting its performer-specific aspect and suggesting an alphabet of prototypical forms for future work. This method is including tempo derivatives in the visualization in addition to the tempo itself. While Grachten et al. use it to visualize expressive tempo modifications, it seems to be highly adequate for visualizing the gestural quality of arpeggiations as well.

A chord struck together defines nothing more than one point in time. An arpeggiation on the other hand defines a region, that can be shaped in a fine-granulated way, as much as a syllable defines a shaped time span. Those regions can be separated from each other by breaks, as the vowels are separated by consonants. The swelling up on stressed syllables is ideally represented by an accordingly shaped arpeggiation, while the swelling down on short syllables tends to be represented by shorter, relatively more vertically aligned chords. In this way, speech and music can be thought as analogous.

Below an example of a place where rhetorical accenting by arpeggiating is almost mandatory:

Thereafter I have tried to apply those rules to my own playing of the chorale. This is what the first try sounded like:

In the video you see the analysis of my own playing. White stems show the mean deviation of the tactus in five different trials to tap to my own performance. The blue line shows the durations in quarter notes – the higer the longer the duration.

Obviously, the goal to "betray" my own tactus was most effective at the "Lass mich […]", where an unexpected hesitation on the upbeat takes place. The principle "short and arpeggiated upbeats lead to accented and arpeggiated downbeats" was most successful towards the end ("nicht gar verzagen").

The layers I have proposed in this research helped me to structurize my thinking about arpeggiation tremendously. In close-listening it is often possible to assign, depending on the context, multiple meanings to one arpeggiated chord. The model of superimposed layers allows to treat those kind of nestings and overlappings very flexible.

Obviously, the layers that I found here only represent a subset of the many more layers that very likely exist. However, to answer the research question – how many different ways of arpeggiation is it possible to think of and how is musical meaning created through them? – my research only shows the number of possible meanings of arpeggiation that I am capable of thinking of at the present state.

Once more it should not remain unmentioned here, that this research does not want to recommend a practise of permanent arpeggiating. Quite the contrary, it seems obvious, that a musically exciting performance can only be achieved by contrasting different interpretative devices. As much meaning can be assigned to arpeggiating, a deliberate non-arpeggiating has its meaning and its right of existence as well. Rather, this work is to be understood as a reflection on the manifoldness of a stylstic device.

Research Question

How many different ways of arpeggiation is it possible to think of and how is musical meaning created through them?


This research aims to discover the expressive possibilities of arpeggiation – may it be indicated by an arpeggio sign or added arbitrarily by a performer. For two reasons this question seems relevant: In the past century, the idea that musical performance should be an exact reproduction of the score has become increasingly predominant and in consequence the idea of exact synchronity has evolved. Seeking for a more creative and less reproducing way to perform, breaking up this rigid synchronity provides us with an exciting "playground" yet to be explored. On the other hand, particularly on the classical guitar arpeggiation can frequently be heard often for more technical than musical reasons. Especially on this instrument, a more reflected use is desirable.
It has barely been looked at the rules on what happens "inside" an arpeggiation. A reason for that might be, that everything inside an arpeggiation usually occurs very rapidly which makes it a lot harder to be analyzed. Besides that, often in arpeggiation the performer relies on "automatic" processes that are difficult to be modified consciously.
I propose the idea, that arpeggiation consists of multiple layers of meaning that can be put together in any possible combination. The layers of arpeggiation span with increasing subtlety from simply habitual arpeggiation to a meticulously planned effect.
I will demonstrate this concept by showing those different ways and meanings of arpeggiation in recordings and on my instrument.
With no claim to completeness, nine of those layers will be proposed and discussed here:

  1. Nonchalance: Contrary to the performance practise preeminent in the 20th century that aims to preserve the integrity of a musical text by reproducing it as faithful and structurally clear as possible, performers from previous centuries did not hesitate to superimpose a layer of improvisatory nonchalance on their playing. This "unrestricted manner" is often expressed by arbitrary arpeggiations all over the place.
  2. Resonance: In order to overcome the missing sustain of a plucked instrument (lute, guitar, harpsichord, etc.) arpeggiation is employed – a practise that can be assumed to be of "longue durée". This practise aims for an "illusion of sustain" – i.e. to make the instrument sound violin-, organ- or even orchestra-like.
  3. Gesture: When an arpeggiation speeds up and slows down in itself once or several times, it can evoke or be part of the complex illusion of an ornamented gesture. Arpeggiations themselves can also be a part of more complex gestures.
  4. Accentuation: With its ability to cover a wide range of different speeds and qualities, arpeggiation can represent any sort of accentuation. An arpeggiated chord can e.g. represent an emphasized syllable of a real or imaginary text underlaying the music. Or, when a musical thought gets interrupted and overlayed by a new one, arpeggiation can represent the rhetorical accent of a new, "sudden idea".
  5. "Legato": Arpeggiations can seamlessly blend from one chord to another, thus creating an illusion similar to the effect of a vocal slide or so-called portamento.
  6. Evocation. An arpeggiation can evoke the illusion of another instrument playing - be it a baroque guitar, a theorbo – even the emulation of multiple instruments creating a closely intertwined sound texture is possible.
  7. Dissolution: A phrase frays out towards its end and through continuous arpeggiation a feeling of instability and disorientation in meter establishes.
  8. Tempo rubato: With multiple instruments involved, precise synchronization of all parts with a given beat has arguably not been sought after for a long time in performance history. Instead of a beat representing a point of time, it seems more appropriate to think of a region, a time span, that is merely approximated by the different parts, while they still are following their own "impetus" independently from each other (commonly known as the earlier type of tempo rubato). Thus, asynchronity is to be understood as the consequence of an exclusively horizontal view.
  9. Suspension: This refers to an effect described by Couperin as "disorienting the ear". It explicitly is used only sparingly by Couperin and is therefore considered here to be fundamentally different from habitual arpeggiation. It refers to an effect of consciously delaying the melody note and thus playing with expectation and imagination of the listener.
  10. Mesh: This effect describes a carefully balanced effect where all the neighbouring notes, vertical as well as horizontal, are considered in the decision of how to dislocate a single note.

From a methodological standpoint, this research is mainly based on the close-listening and analysis mostly of historical, but partly also of modern sound recordings. In some chapters, written historical sources have been read. Words are not suitable for describing how exactly a specific kind of arpeggiation works – it is not far-fetched to assume that the "art" of arpeggiation was primarily aurally taught. Therefore, written sources are often not particularly exhaustive. Visualization, imitation and experimentation of remarkable sound documents has led to the following results.


In this research different examples will be presented, some played on the guitar, some on the harpsichord. If not stated otherwise, they are all played by myself.

In order to give an idea of the outcome of this research I first present a phrase of a Sarabande by Francois Couperin, L'unique, played on the harpsichord. I anatomized it into the above-mentioned layers. In the recording I play the whole Sarabande, while the analysis includes only the first eight bars. In the score below, the graphical size of a bar is proportional to the duration of that particular bar in the recording.

After around 20 attempts to adapt myself to the mentioned traits of Llobet's playing – expanded and slowly arpeggiated or dislocated first beats, shortened and quickly arpeggiated second beats, a hand relaxedly and rapidly "falling" into the strings – the result sounds like this:

This style of arpeggiation as Llobet employs it, can be characterized as "gesture" that superimposes a layer of nonchalance and improvisation on the actual composition.

Below another example of three successive chords from Chopin's op. 28 No. 7 that tries to transport a feeling on nonchalance. In the following chapters I am going to come back to this particular piece.

The graphs should be read as trajectories. While the horizontal axis represents tempo itself in beats per minute (BPM), the vertical axis represents the change (derivation) of tempo (in mathematical terms accelaration). Every single note of the arpeggio is represented by a point that is connected with the point of the next note. The music starts at the thinner end and ends at the thicker end of the trajectory.

I will further demonstrate this form of visualization by comparing four different interpretations of harpsichordists playing the opening arpeggio of the Prelude nonmesuré "à l'imitation de Froberger" by Louis Couperin.


In singing we can observe a phenomenon that occurs all over vocal and instrumental early recordings: cercar della nota and the portamento. Especially the frequent use portamenti was a required feature of a vocal or vocal-like performance style in the 19th century. The primary goal of portamento is the seamless connection of notes, which ultimately serves the idea of legato. The most plausible devices to imitate on the piano a singing style, that indulges in the sound of portamenti, are dislocation and arpeggiation. In my opinion it is therefore, that Carl Czerny, who positions himself critical to the common habit of  his time to arpeggiate all over the place, expressly allows arpeggiation of all chords in phrases that are singing-like ("Accorde, welche einen Gesang bilden").

For a sounding demonstration of how the ideal of a "real" legato through the permanent use of portamenti might be resembled by arpeggiation, I take an excerpt of the Aria "Ah non credea mirarti" (Bellini, Somnambula) as sung by Adelina Patti.1

tempo rubato

Contrasting to the previously mentioned kinds of arpeggiation, the effect of "tempo rubato" has primarily horizontal causes. Of particular interest is here what Richard Hudson in his study on the history of tempo rubato calls the "earlier type" of tempo rubato: A melody follows it's own impetus while the accompaniment is keeping a more or less steady tempo – thus the melody anticipates or lags behind the accompaniment and thus the effect of an always changing, often non-linear arpeggio is created.

In the above example you can here a rather extreme kind of deliberate and idiosyncratic tempo rubato. It certainly shows to which extend synchronity was obviously not sought after. At most we find a rough approximation of it. While in this particular case a quite regular melody-lead can be observed, all the different orders (in total six) can possibly appear – the piano as first, second or third instrument, likewise the violin and the cello.

This recording was done in 1905 on a Welte-Mignon reproducing piano roll. Those piano rolls accurately reproduce the timing of the performers while the information on dynamics are less reliable. There has been a sufficient discussion on the accuracy of piano rolls in the literature already and there is no need to repeat it here.1

In this context, a remarkable statement by Saint-Saens about his way of playing rubato should not remain unquoted:


"Through Mme. Viardot [pianist, singer, one of Chopin's closest associates during the last decade of his life] [...] I learned the true secret of tempo rubato [...where] the accompaniment holds its rhythm undisturbed while the melody wavers capriciously, rushes or lingers, sooner or later to fall back upon its axis. This way of playing is very difficult since it requires complete independence of the two hands; and those lacking this give themselves and others the illusion of it by playing the melody in time and dislocating the accompaniment so that it falls beside the beat; or else - worst of all - content themselves with simply playing one hand after the other. It would be a hundred times better just to play in time, with both hands together." – from Quelques mots sur l'éxecution des oeuvres de Chopin, 1910.


It should further be noted, that Saint-Saens finally uses arpeggiation regularly to create a strong forward motion: often arpeggiation occurs on the third and fourth beat and hence give the listener a sensation of being "pulled" into the next first beat.

In the following experiment I have tried to apply the found priniples to my own playing – in this case to a phrase from Schubert's song Liebesbothschaft, in a 19th-century transcription by Johann Kaspar Mertz.

Which sounds as follows (played twice: first in a strong rubato-manner, subsequently in a more synchronized way for comparison):

F. Couperin, L'enfantine from the Seconde livre,bar 3f.


In the chapters above I tried to show how "simple", linear arpeggiation can be used, to make the instrument sound more resonant, to play legato, to accentuate and to play in a tempo rubato style. Arpeggiation can further be used as an evocative effect, in order to "awake" the listener's memory of a particular instrument or a particular idiomatic performance style. I will give two demonstrations of what that can sound like. In the following examples we encounter a complex form of arpeggiation, that does not go simply up- or down, but allows nested changes of directions, internal repetitions, passing notes etc.

For example, a particular way of arpeggiation can make the classical guitar actually sound similiar to the strumming effect on a baroque guitar:


Reduced to a basic principle we could state that in order to imitate the sound of a reentrant tuned instrument a small group of notes is repeated before the arpeggio continues. Louis Couperin probably had this idea in mind in his opening arpeggio for the "Prelude a l'imitation de Froberger":

In the text below the proposed layers will be discussed further. I am going to refer back to the fragments of this recording, explaining why they have been assigned to a specific layer.

The goal of this research is to open up to performers an incredible wide field of possibilities for interpretation and experimentation. It does not want to suggest that everything should always be played in an arpeggiando-manner.

The quality of the arpeggi of Llobet very often leaves the impression, that they are based on a relaxed and "free" right hand (and thus often erraticly accented) that allowes the hand to drop into the strings with a wrist bent to around 90 degrees. The usually uneven internal timing of the fast arpeggi suggests, that Llobet often employs a thumb that in a rapid, "falling" gesture strucks multiple strings in a quick succession.


The primary aim of  arpeggiation is to make the instrument sound more full and resonant than it actually is – to make the sound in a way bloom. This superimposes a layer of improvisatory sprezzatura, which has most likely been a standard feature of musical performance until the 20th century, when the concepts of reproduction and authenticity replaced the idea of improvisation.

Point of reference for this phenomenon is the instrument itself. The need to strike notes again or to spread chords is founded in a specific characteristic of the instrument itself. Nothing else is meant by Frescobaldi when he instructed in his preface to the toccatas to "not leave the instrument empty" ("non lasciar vuoto lo strumento"):

"Li cominciamenti delle toccate siano fatte adagio, et arpeggiando; e così nelle ligature, o uero durezze, come anche nel mezzo del opera si batteranno insieme, per non lasciar uoto l'Istromento; il quale battimento ripiglierassi a bene placito di chi suona."

"The beginnings of the toccatas should be played slowly and arpeggiated. In suspensions or dissonances, as well as in the middle of the work, [the notes] should be struck together in order not to leave the instrument empty; and this striking might be repeated as the player likes."

Tagliavini points out1:

"Given the brevity of sound of quilled instruments, one must strike the held notes again in order to maintain the effect and flavour of the dissonance. This technique should be used not only for dissonances, but also elsewhere in the course of a piece (’in the middle of the work’) to keep the sound of the instrument alive and full. Frescobaldi’s aim was, as Diruta puts it, to ’produce the same effect of holding the harmony on the quilled instrument as the air does in the organ’".

In a similiar way the entry in Rousseau's Dictionaire de musique on the arpége 
is to be understood:2

"What is done on the violin by necessity, it is practiced by taste on the harpsichord. Since we can only get from this instrument dry sounds that do not hold, we have to rewind them on long notes. To make a chord last longer, it is struck by arpeggiation, starting with the low sounds, and observing that the fingers which have struck the first must not leave their touch until all the arpeggio is finished, so that we can hear all the sounds of the chord at once."

Anselm Gerhard, who collected numerous evidence of habitual arpeggiation and stated its "longue dureé", mentions a later German source, Jakob Adlung, stating:3

„Das Spielen auf solchen beseyteten Instrumenten [Clavicymbel, Clavicytherium, Spinet] ist anders, als auf der Orgel. Man muß sich mehr der Brechungen und dergleichen befleißigen, als daß man die Claves zusammen oder allzulangsam anschlägt; denn die Seyten hören bald auf zu klingen.“

["The playing on stringed instruments [clavicymbel, clavicytherium, spinet] is different than on the organ. One has to make more use of arpeggiations and the kind, than striking the keys together or too slowly, for the strings soon stop sounding."]

The necessity of arpeggiation in order to keep the instrument sounding is more significant for historic instruments, that have a less brilliant and sustaining, but more transparent sound than their modern equivalents.4 Breaking and arpeggiating is not only a habit on quilled keyboard instruments. Their permanent use can also be considered to be the characteristic trait of lute playing at the time and it is probably what Francois Couperin meant, when he was mentioning ”les choses luthée” and certainly what later Johann Gottfried Walther meant when specifying the ”Lauten-Art” (manner of the lute) as ”arpeggiando oder gebrochen”.4 Perrine writes:5

"The particular manner of playing all kinds of pieces for the lute consists only in arpeggiation or the separation of voices, as I have notated in the majority of lute pieces put into musical notation below."

As the device is idiomatic for the lute as well as for the harpsichord we can safely assume that frequent arpeggiation and separation were on both instruments applied without further indication, left to the ”discretion” of the player. Regarding its musical character, another English source mentions that this ”art” defines the ”soul” of the lute and makes it sound ”airy and skipping”. It is usually agreed, that the origin of this habit on both, harpsichord and lute, is to be found in the role of the mentioned instruments for accompaniment.
Since it is a device, that so easily allows us to make the instrument sound more rich and full, it obviously became and remained a habit, a ”manner”, on plucked instruments as well as on keyboard instruments. Jumping forward in time, we find e.g. Carl Czerny complaining:

"Manche Spieler gewöhnen sich das Arpeggieren der Accorde so sehr an, dass sie gar nicht im Stande sind, vollgriffige Accorde, oder auch nur Doppelnoten, vollkommen fest und auf einmal anzuschlagen".
["Some players get so much used to arpeggiating the chords that they are not able to strike full chords or even just two notes firmly and at once."]

This is a fascinating document of Carl Reinecke representing a performance style that is still rooted not only in the tradition of improvising or playing improvisation-like, but also preserving an arpeggiated manner that goes back to the century-old habit that fights the missing sustain of keyboard instruments.1 Especially valuable are the rolls for the Welte-Mignon system, as they preserve the exact timing in a most authentic way, but also rolls of the Hupfeld studio in Leipzig are worth being noted. Carl Reinecke, born 1824, praised already in 1835 as a "gracious Mozart player" and being a teacher in the rather conservative Leipziger Konservatorium,  presumably preserves a traditional style of playing – perhaps even of the 18th century.

Regarding Reinecke's obviously regular use of arpeggiation, Daniel Leech-Wilkinson already raised the question, if the practise of arpeggiation in connection with the habit of delaying the melody note, might be pointing us to a much older tradition:2

"Most chords are arpeggiated upwards, so consistently that one wonders whether this is harpsichord technique surviving into nineteenth-century playing, or nineteenth-century pianism applied to Mozart. The notes of the melody are almost always delayed, sometimes by as much as 1/5 of a second, which may not sound very much but, when one is used to notes placed vertically in a score being played exactly together, seems a very long time as one listens. […] In Reinecke’s case, the regularity of the arpeggiation and of the accompaniment beat lead us strongly to the sense that the melody is late rather than the bass early, and that too tends to increase the sense of alienation for a modern listener, who is at least familiar with the idea of a low bass note sounding early when an impossible stretch forces it. Reinecke’s playing can’t be explained away like that, and the fact that he is playing Mozart, for us such a regular composer, simply makes the strangeness of his playing more acute. The thought that there might be a historical cause in his youth is almost frightening because of the wholesale rethink it would force about everything we imagine as Classical."


From those visualizations, some conclusions can immediately be drawn: Magdalena Hasibeder breaks the arpeggio down into three gestures, while every gesture is faster than the previous one, and finishes with a strong retard. Francesco Corti on the other hand interprets the arpeggio as only one gesture by very evenly speeding up and slowing down. His interpretation represents the rapid gesture of all four interpretations. Gustav Leonhardt starts with a very fast arpeggiation, gives more time subsequently in order to initiate two gestures from there. He finishes rather fast.

In the above presented Sarabande L'unique by Couperin, gestural arpeggiation happens very frequently. The graphic below shows every note attack represented as a vertical line on a time axis. The more lines are densely packed together, the more note attacks are occuring in that place. This visual representation allows to easily group together a set of notes as one gesture.

Together with the notes of the first beat of bar 2, bar 1 arguably builds one gesture composed of several smaller gestures. In the proposed method this gesture could be visualized like this (the labels should be read as bar/beat):


Arpeggiation covers the whole palette from strength to softness, from determined to indefinite and thus is the perfect mean to subtly accent certain chords, spanning from stressed to unstressed with all the innumerable shades in between.

It has often been claimed, that the musical accentuation follows the accent-structure of language. While the claim to play speech-like, rhetorically, is not a particular new one, it has to my knowledge never been explored what it sounds like, if the sound of our playing resembles as closely as possible the sound of a spoken declamation.

Starting point is the observation that stressed syllables are  naturally "shaped" in a characteristic way by a melodical swelling up and down of the voice. It should be noted, that this observation is by far not a new one. As early as 1779, Joshua Steele invented in his Prosidia naturalis a notation that precisely shows this phenomenon in the English language (p. 13):


In the already introduced recording of L'unique by Couperin, there is a place in the repetition of the first part that can be sensed to represent a rhetorically accentuated arpeggiation. This arpeggiation takes place on the second beat of the first bar:

In order to experiment with the hypothesis, that the effect of emphasized syllables can be resembled by arpeggiation, I have been conducting the following experiment: First, I recorded myself speaking the underlying text of a chorale by J. S. Bach in a more or less declamatory way. Afterwards, I aligned in a MIDI editor the notes of the chorale precisely to the spoken version. In the case of melodically blooming syllables an arpeggiation has been added that takes the exact amount of time as the syllable does:


Some general rules can be deduced from here:

  1. Upbeats are shortened and connected with the following main accent by arpeggiation, which is arpeggiated itself again.
  2. Unemphasized ending syllables are struck together.
  3. Permanent tempo inflection is a quality inherent to the sound of speech. The first two rules are only effective if the license is taken to quantitatively add time or take away time from the length of a note.

It is interesting to note that a little survey has shown the tendency that non-musicians have less difficulties accepeting this heavily arpeggiated and inflected version, while musicians usually prefer the more stable version. Applying this principle to my playing has proven to be extremely difficult. This little recording shows a deliberate idiosyncrasy in timing and arpeggiation that comes in my opinion very close to the style heard in the early 20th century recordings.

When one arpeggiation or dislocation is immediately followed by another one, soon a feeling of fragility and instability establishes. I call this phenomenon „dissolution“. Used sparingly, this device can be very effective. A listener can easily recognize this device when he is unable tap the meter. Meter usually requires a certain stability that in this case is threatened by a kind of arpeggiation that deliberately introduces local instability and disbalance.

Having listened to this example several times and having accustomed myself to this style of playing, the following heavily desynchronized performance of Saint-Saens playing Chopin's op. 15 No. 2 seems less incomprehensible than it might haven seemed when listening to it "unprepared". Since I have shown, that in ensemble playing still as late as in 1929 we can hear performances that are deliberate out-of-synchronization, it seems plausible to argue that Saint-Saens is trying to play as if two performers were playing together.

The leap in bar 6 could be an example for a place where Saint-Saens tries to create the illusion of a vocal portamento (see the section on Legato above).

What happened?

  • agogical lengthening of beginning, middle and end of the phrase
  • two times melody before accompaniment, otherwise always melody after. Only the final phrase note played together with accompaniment
  • The accompaniment 32nd-note figure seems fairly regular in timing, except for the 2nd and 3rd beat in bar 1, which is at the same time most dense and "exciting" place


In 1717, Couperin describes the two agogic ornaments aspiration and suspension (on which we focus here) as follows:

"The emotional response, to which I refer, is the result of a deliberate cutting short or delaying of notes in a manner appropriate to the manner of each particular prelude or piece. By opposite means these two ornaments disorient the ear. In the same way that string instruments swell the tone, delaying notes on the harpsichord gives the ear the impression of a similiar effect."3


Couperin writes his sign for the suspension only sporadically and in varying contextes – in cadences (L'Enfantine), single bars that seem like taken out of time (Les Delices) or in a melodic context that gets a hesitating character through the suspension (La Castelane).

Considering the places in which Couperin uses the suspension, it seems to be a natural conclusion to see the suspension as an open space that has to be filled by the performer with something imaginary. Thus it is more than just an arpeggiation in order to make the instrument sound fuller. It has a rhethorical function. This imagined space could be for example the bloom of a violin tone, but also a appogiatura from up or down seems possible.

Another common device for arpeggiation on baroque guitar as well as theorbo is to pull backwards the index finger through multiple strings. Combined with the effect of imitating a reentrant tuning, the same arpeggio of the above sound example is as follows:

The same principle is also applicable for less extended arpeggi:

The Illusion of Multiple Players

To attract the listener's attention, arpeggiation has to sound improvised and multi-dimensional – as if it was being improvised by two or more persons playing together (e.g. basso continuo) and the outcome is not just convincing, but excitingly complex.

In order to demonstrate what is meant by that kind of "complex arpeggiation", I will give as an example a recording of a violin sonata of Händel (HWV 371, movement Affetuoso) played by Julia Schröder, Daniele Caminiti and Giorgio Paronuzzi. The goal is to imitate these multi-layered arpeggi of the basso continuo (harpsichord and lute) on the guitar. In a first step, the arpeggio-timing is being visualized using the Software Sonic Visualizer (which I also used for the visualizations in the above graphics and videos). Red lines represent the harpsichord, green lines the lute. In a second step I try to "transfer" a specific arpeggio to the guitar.

Applying Multi-Layered Arpeggiation Timing to a Notated Arpeggio

I assume that the ideal timing for an arpeggio is to be found in the improvisation of two players who react on each other. For experimentation we imagine a notated arpeggiation like in this case the opening bar of the Allemande BWV 825 by Johann Sebastian Bach to actually represent the idea of two improvised arpeggi, that are correlating with each other.

What did happen when simulating the arpeggio of this bar to be actually played be two different persons?

  • strong agogics (accelerando to the middle and a strong ritardando to the end)
  • due to the free ("improvised") timing a metrical accent can only be felt on the first beat
  • in the "syncopated" place the upper voice has a slightly slower tempo, thus changing its position from being precisely between two notes more becoming a upbeat to the next note
  • the syncopated notes are played slightly stronger then the non-syncopated notes which results in the feeling of a subtile "anti-metrical" accent
In the L'unique recording this "interlocking" of two arpeggiations can be heard in the second part of the piece:


I introduced this research with a recording of a piece by Francois Couperin played on the harpsichord and I finish it with a recording of a Prélude by Frederic Chopin on the guitar. It is the Prélude No. 7 op. 28 transcribed for guitar by Francisco Tárrega. I have introduced fragments of it in the above chapters already.


  1. [1]  David J. Buch. “”Style brisée, Style luthée,” and the ”Choses luthées””. In: The Musical Quarterly 71.1 (1985), pp. 52–67. issn: 00274631, 17418399.

  2. [2]  Robert Hill, "Carl Reinecke's Performance of Mozart's Larghetto and the Nineteenth-Century Practise of Quantitative Accentuation", in: Gregory G. Butler, ed. About Bach. Urbana, Ill.: Univ. of Illinois Press, 2008. isbn: 0-252-03344-2; 978-0-252-03344-5.

  3. [3] Jesper Bøje Christensen, "Was kein Notentext hätte erzählen können. Zur musikalischen Bedeutung und Aussagekraft historischer Tondokumente", in: Claudio Bacciagaluppi, Roman Brotbeck, and Anselm Gerhard. Zwischen schöpferischer Individualität und künstlerischer Selbstverleugnung: zur musikalischen Aufführungspraxis im 19. Jahrhundert. Vol. 2. Edition Argus, 2009.
  4. [4]  Philip Corri. L’anima di musica, an original treatise upon pianoforte playing, etc. London, 1810.

  5. [5]  Francesco Corti. Suites for harpsichord. Genuin.

  6. [6]  Carl Czerny. Vollständige theoretisch-practische Pianoforte-Schule. Wien:

    Diabelli, 1846.

  7. [7]  Girolamo Frescobaldi. Orgel- und Clavierwerke. Ed. by Christopher Stembridge and Kenneth Gilbert. Urtext. Vol. 1,2: Toccate e partite dintavolatura di cimbalo ... libro primo (Rom, Borboni, 1615, 1616). Bärenreiter Urtext. Kassel: Bärenreiter-Verl., c 2010.

  8. [8]  Anselm Gerhard. “You do it! Weitere Belege für das willkürliche Arpeggieren in der klassisch-romantischen Klaviermusik”. In: Zwischen schöpferischer Individualität und künstlerischer Selbstverleugnung. Ed. by Claudio Bacciagaluppi, Roman Brotbeck, and Anselm Gerhard. Musikforschung der Hochschule der Künste
    Bern. Schliengen: Edition Argus, 2009.

  9. [9]  Hermann Gottschewski, "Tempoarchitektur – Ansätze zu einer speziellen Tempotheorie", Musiktheorie, 8(2), pp. 99–117 .
  10. [10]  Hermann Gottschewski. Die Interpretation als Kunstwerk. Musikalis- che Zeitgestaltung und ihre Analyse am Beispiel von Welte-Mignon- Klavieraufnahmen aus dem Jahre 1905. Vol. 5. Freiburger Beiträge zur Musikwissenschaft. Laaber: Laaber, 1996. isbn: 3-89007-309-3.

  11. [11]  Maarten Grachten et al. “Phase-plane Representation and Visualization of Gestural Structure in Exprssive Timing.” EN. In: Journal of New Music Research (2009).

  12. [12]  Maarten Grachten et al. “Phase-plane Visualizations of Gestural Structure in Expressive Timing.” EN. In: In Proceedings of the Fourth Conference on Interdisciplinary Musicology (CIM08), Thessaloniki. July, 2008. 2008.

  13. [13]  Magdalena Hasibeder. Frobergers Reisen. Edition Raumklang.

  1. [14]  Bruce Haynes. The pathetick musician : moving an audience in the age of eloquence. Ed. by Geoffrey Burgess. New York: Oxford University Press, [2016]. isbn: 978-0-19-937373-4.

  2. [15]  Daniel Leech-Wilkinson. 2009. url: chapters/intro.html (visited on 02/28/2019).

  3. [16]  Gustav Leonhardt. Suites and Pavane. In: Editio classica BMG classics. Deutsche Harmonia Mundi.

  4. [17] Johannes Monno, Die Barockgitarre. Darstellung ihrer Entwicklung und Spielweise, München: Tree Edition, 1995.
  5. [18]  Adelina Patti. Ah! non credea mirarti : La Sonnambula. before 1915.

  6. [19]  Neal Peres da Costa. Off the record : performing practices in romantic piano playing. New York: Oxford University Press, 2012.

  7. [20]  Julia Schröder et al. Händel: Violinsonaten. Deutsche Harmonia Mundi, 88697885782.

  8. [21]  Skip Sempé. Louis Couperin. In: Ut pictura musica. Alpha.

  9. [22]  Sigurd Slåttebrekk and Tony Harrison. Chasing the Butterfly. 2008f. url: (visited on 02/28/2019).

  10. [23]  Luigi Ferdinando Tagliavini. “L’Arte di non lasciar vuoto lo strumento: Appunti sulla prassi cembalistica italiana nel Cinque e Seicento”. In: Rivista Italiana di Musicologia 10 (1975).

  11. [24]  Luigi Ferdinando Tagliavini. “The Art of ’Not Leaving the Instrument Empty’: Comments on Early Italian Harpsichord Playing”. In: Early Music 11.3 (1983), pp. 299–308. issn: 03061078, 17417260. url: http: //

  12. [25]  . The Guitar Recordings ; 1925 - 29 : music by Aguirre, Albéniz, Bach, Coste, Llobet, Mendelssohn, Ponce, Sor, Quijano, Villar. Chanterelle.

  13. [26]  Claudia de Vries. Die Pianistin Clara Wieck-Schumann. Mainz, 1996.



While mainly due to the development in the field of recording technology an aesthetic of precision and control emerged in the 20th century and therefore the development of instrumental playing is commonly thought of as a development of constant progress and improvement, there is a pessimistic perspective to the optimistic one – a perspective, that would rather describe the 20th century as a history of lost spontaneity, improvisation, nonchalance.

The stylistic device – or, from the other side, "vulgarity" – of arpeggiation fell first victim to this development. Edward Elgar recorded in 1929 five piano improvisations. They show an idiosyncratic playing that is oriented on an orchestral sound and is in a way "careless" and unrestricted in terms of synchronity.

In this place something occurs, which might be called flattening of the hierarchy between melody and accompaniment. From the perspective of timing they are for a short moment one unit. This effect seems so beautiful because is dangerously close to chaos.

 Multiple reasons lead to the decision to rather extremely dislocate the melody note e in bar 2:

  • a singer educated in a 19th century performance style would quite certainly have embellished it with a portamento
  • it represents the local registral extreme
Another example of applying tempo rubato on the guitar (Chopin Prelude op. 28 No. 7):

Another idea is the imitation of the reentrant tuning of a theorbo. Prominent feature is the non-linearity of arpeggiation as a result of this reentrant tuning.

The Art of Arpeggiation

and Other Asynchronities

This recording is remarkable for its unfiltered transportation of the players musical intentions beyond any technical considerations or pianistic restrictions.

I will try here to get a certain kind of "inside perspective" on an early, heavily arpeggiated guitar recording.1 It is a transcription of a folk song, which Miguel Llobet "harmonized" in 1900. The recording of Llobet playing himself was done in 1925. A first listening to this recording already shows a huge variety of different arpeggio lengthes. This is particularly noticeable in comparison to a modern performance of the same piece.

This video shows the performance of Miguel Llobet. The longer a stem, the more time the arpeggio takes.


Especially for the last chords of the fast movement of this Froberger partita the pulled index seems to create an interesting effect referencing the sound of a baroque guitar:

Obviously the aim of the two portamenti in bar 1 (upwards) and bar 2 (downwards) is to seemlessly connect two notes by sliding.1 This is what I seek for, when I try to imitate the portamento-style of Patti on the guitar with arpeggiation: to blend one chord into the other by arpeggiating them "delicately" and by playing often before the beat in order to "blur" the line of exact separation between the chords.

A pianist (or guitarist) who has this vocal style as a model will naturally employ a dislocated and asynchronous way of playing. While on the piano the time that is filled by the portamento remains imaginary, it is in some cases possible to imitate it on the guitar. I therefore suggest the use of dislocation to imitate this bel canto style on a keyboard instrument while on the guitar a combination of portamento and dislocation resembles it most closely:


Violin is neither part of the visualization nor the transcription. Played in 415Hz

In the example of  my recording of Couperin's Sarabande, the same place that represents a dissolution also includes a suspension. It is the e' in the right hand on the second beat (to be read as an alto clef), that occurs later than the bass note and has a characteristic break ("emptiness") before it sounds. It deliberately "betrays" the listener's expectation:

This is an example for dissolution from the above-presented recording of a Sarabande by Francois Couperin. Towards the end of the bar the feeling of an ordered meter is more and more fading, a feeling of unpredictability takes its place.


A picture of Miguel Llobet clearly showing his right hand position. This position was preferred by Francesco Tarréga, Llobet's teacher. Andrés Segovia also used it. 

This is the opening bar of a Sarabande by Johann Jacob Froberger, transcribed for guitar by me. The effect of playing the opening chord in a way like the strumming on a baroque guitar might be most suitable.

Right-hand fingering: d–a–m–i–p. Very slowly strum the strings and follow with the next finger before the previous one has finished. This technique of strumming was also used on the baroque guitar.1

Another example from Chopin's Prelude op. 28 No. 7 (arranged by Francesco Tarregá). The effect of connecting the upbeat with the following downbeat in bar 1 is reached by placing the accompanying a slightly before the downbeat.

The suspension thought as a non-sounding appogiatura

transformed (and for technical reasons transposed to D major) to a hypothetical version for two independent players:

This is a graphical representation of my own playing. Green stems show the duration of 32nd notes in the accompaniment, the red line shows the duration of eighth notes in the melody. Text boxes show the counting in eighth notes.

Text: "Rauschendes Bächlein, / so silbern und hell"



of the basso continuo, played in 440Hz

The recordings are played in the following order: (1) Hasibeder, (2) Francesco Corti, (3) Skip Sempé, (4) Gustav Leonhardt

Girolamo Frescobaldi, Toccate e partite d'Intavolatura, Rome: Niccolo Borbone, 1616. Translated by Christopher Sternbridge and Kenneth Gilbert (edd.), Kassel: Bärenreiter 2010.

6 David J. Buch, "'Style brisée', 'Style luthée and 'Choses luthées", MQ Vol. 71 No. 1 (1985), p. 57f.

This recording has been studied already various times, mostly regarding Reineckes significant use of tempo modification. See e.g. Robert Hill, "Carl Reinecke's Performance of Mozart's Larghetto and the Nineteenth-Century Practise of Quantitative Accentuation", in: G. Butler, G. Stauffer, M. Greer (edd.), About Bach,  Urbana 2008, pp. 171–180 and Hermann Gottschewski, "Tempoarchitektur – Ansätze zu einer speziellen Tempotheorie", Musiktheorie, 8(2), pp. 99–117 .


2 Daniel Leech-Wilkinson, The Changing Sound of Music: Approaches to Studying Recorded Musical Performance, Ch. 6 ¶24.

1 cf. Johannes Monno, Die Barockgitarre: Darstellung ihrer Entwicklung und Spielweise, München 1995.

1Luigi Tagliavini,  "L'Arte di «non lasciar vuoto lo strumento»: Appunti sulla prassi cembalistica italiana nel Cinque e Seicento", Rivista Italiana di Musicologia Vol. 10 (1975), p. 375. see also Diruta Il Transilvano I, f. 6: “[…] quello stesso effetto, che fa il fiato nel Organo, nel tener l’armonia, bisogna che fate fare all’Istrumento da penna; Et per eßempio, quando sonate nel Organo vna Breue, over Semibreve non si sente tutta la sua armonia senza percuotere più d’una volta il Tasto: ma quando sonarete nell’ Istrumento da penna tal note li mancherà più della metà dell’armonia: bisogna dunque con la viuacità, & destrezza della mano supplire à tal mancamento con percuotere più volte il Tasto leggiadramente.”

2Jean-Jacques Rousseau, "Volume 9. Dictionnaire de musique", in: Collection complète des oeuvres, Genève, 1780-1789, vol. 9, in-4°, online edition at, last accessed 3.3.2019. p. 27.

3 Anselm Gerhard, "'You do it!'. Weitere Belege für willkürliches Arpeggieren in der klassisch-romantischen Klaviermusik", in: Zwischen schöpferischer Individualität und künstlerischer Selbstverleugnung.Zur musikalischen Aufführungspraxis im 19. Jahrhundert, ed. by Claudio Bacciagaluppi, Roman Brotbeck und Anselm Gerhard, Schliengen 2009 (Musikforschung der Hochschule der KünsteBern,Bd.2),S.159–168, here: p. 160

4 A typical comparison of the construction of modern pianos vs. historical fortpianos and the implications for the interpretation can be found in Clauda de Vries, Die Pianistin Clara Wieck-Schumann. Mainz 1996, p. 59–67. Unfortunately she describes the development of keyboard instruments as a history of constant improvement which might be reconsidered. Similiar observations can be made for the development of the classical guitar.

7 Perrine, Pieces de Luthé en musique, Paris 1680, p. 5, cited by David J. Buch, "'Style brisée', 'Style luthée and 'Choses luthées", MQ Vol. 71 No. 1 (1985), p. 58.

Carl Czerny, Pianoforte-Schule 3. Theil, op. 500, p. 40.

This is the second movement of the Krönungskonzert by Mozart (listen to it here)

These graphics as well as all the following ones, are done by myself using mainly two tools: The software Sonic Visualizer, which allows to precisely analyse the timing of recordings, and the programming environment R Studio, which I subsequently used for the visualizations.

The left graph shows an extract of the spectrogram of Patti singing. The vertical axis shows the frequencies (from wich the pitch becomes obvious) with all it's fluctuations and the horizontal axis shows time.

This is a recording of the "London Derry", played by David de Groot, violin, D. Bor, piano, H. M. Calvé, violoncello. Recorded in 1929. Found here.

I have been conducting an experiment in which I tried to play with all the possible permutations of the notes of a chord. Though I would consider this experiment to be failed, it is documented here.

from the "Huitième Ordre" of the Second Livre de Pièces de Clavecin, Paris 1716.

1 Already Domenico Corri describes this to be the main point of the portamento (The Singer's Preceptor, London ca. 1810, p. 3f.): "Portamento di voce is the perfection of vocal music; it consists in the swell and dying of the voice, the sliding and blending one note into another with delicacy and expression". For further discussion see Jesper Bøje Christensen, "Was uns kein Notentext hätte erzählen können: Zur musikalischen Bedeutung und Aussagekraft historischer Tondukumente", in: Claudio Bacciagaluppi et al. (edd), Zwischen schöpferischer Individualität und künstlerischer Selbstverleugnung: zur musikalischen Aufführungspraxis im 19. Jahrhundert. Vol. 2. Edition Argus, 2009. p. 146.

1 the most in-depth study of this topic has been done by Hermann Gottschewski, Interpreation als Kunstwerk, Laaber 1996. Also Neal Peres da Costa, Off the Record, Oxford 2012 thoroughly discusses the question of piano rolls.

3 See Bruce Haynes, Geoffrey Burgess, The Pathetick Musician. Moving an Audience in the Age of Eloquence, …, p. 226. Rameau uses the same sign as Couperin.

opening arpeggio of Johann Jacob Froberger, Sarabande from Partita in a, FbWV 615.

1 In their project "Chasing the butterfly", Sigurd Slåttebrekk and Tony Harrison have undertaken the endeavor to "recreate" the 1903 recordings of Edward Grieg by rigorously internalising Grieg's playing, by "understanding his performance style from the inside" and therefore "empowering" themselves to "demonstrate what these performances would have sounded like today had they been recorded with modern recording equipment".1Slatterbrekk and Harrison have thoroughly documented this project on the wbsite .

= Couperin's sign for a suspension

(or recorded in only one track:)

sounds like this: (recorded in two different and later synchronized tracks)