“The notes must embrace the bars, not the bars the notes” – Robert Schumann

In pursuing this project, we have wanted to single out, describe, and demonstrate some important characteristics of Grieg´s playing. Other performers of his time share these characteristics, many of which have gradually disappeared from performances in our own time. The choice of examples we have made throughout this web is of course made to support our case. It would be quite possible to ‘prove’ or demonstrate examples in opposition to this, both in modern and in early performances. There are exceptions in both directions, most certainly when we concentrate on single aspects of a performance. These are by no means black and white issues! But even though many performers and theorists, most specifically from the early music scene, have contributed immensely to our understanding of tempo modulation and tempo modification, there are in our view still aspects of the late romantic performances that most performers of today do not fully comprehend, specifically with regards to phrase junctions, musical syntax, and phrase structure. Although there is good reason to believe that many of the performance features we find in the early years of the recording era originated in even earlier performance styles, we have chosen to present ‘our’ Romantic performers in their performances of the Romantic repertoire. In this repertoire we often find a prevailing cantilena ideal, and the performed melodic idioms correlate closely with the composed idioms.


When asked about the term ‘musical narrative’, I have often referred to a speech once given by my then two-year old son. He was very much of a mummy´s boy, and when sensing there was an argument building up between his parents, he firmly entered the arena. He had brought his sword along, was gesturing vividly, and performed a very convincing spoken defence of his mother. There was indeed a clear structure to the argumentation, there was even some kind of syntax, and the performance was highly emotional. His communication wasn´t any less persuasive from the fact that not one single word was recognisable. His point was perfectly clear: ”You stay away from MY mother!”

In a very similar way, music sometimes displays the most convincing narrative qualities without the support of a single word. This is of course not to say that music is mainly a substitute for words, suitable only for those of us not capable of fully expressing ourselves verbally. The beauty of music is the way it communicates the added content. In the best of performances, music is often carried forward on a narrative structure, but with its specific content resonating with quite different parts of our consciousness than we use when decoding a verbal message. There is certainly a gestural component to this as well, and music´s ability of momentarily articulating our most intimate feelings such as despondency, joy, jealousy, grief, fear, and love, is often the result of a very delicate blend of the gestural and the narrative content. Without the verbal logic, and perhaps specifically for the lack of verbal logic, music has the magical ability of articulating the unspeakable.

A very large part of our music tradition rests on the rhetorical principles of the ancient Greeks and the nature of narration. As early as 1739 Johann Mattheson commented sarcastically in ‘Der vollkommene Capellmeister’ on a contemporary, ‘innovative’ poet pointing out the close relationship between the art of rhetoric and music. “Some years ago a great German poet thought he had made the unique discovery that music is almost exactly the same in this regard (the theory of incisions) as rhetoric. How amazing!

Without dwelling here on the tracing of specific rhetorical models in written music and various performances, it is certainly worth considering just how the performer needs to contribute if the communication of the narrative content is to make any sense. What are our ‘tools’, and how do we use them in our communication? In the same way that we previously defined some of Grieg´s ‘inverse characteristics’, it is also worth considering which elements are capable of delineating or diminishing the narrative qualities. Could there be elements of a performance that hinder the natural narrative, this being such an essential part of romantic music-making?


As classical performers, we normally operate within a rather tight zone. Performers from other genres are in fact often shocked by the apparent and actual limitations we are dealing with when performing historical music. Naturally, there is a lot to be said about this, and most of us do not fully explore the interpretational possibilities within our own genre – most specifically in the domain of improvisation. In the romantic repertoire there is however less room for improvisation than in earlier music. The performer´s operational zone is seems to be even further reduced. In the Romantic period we are usually presented with a highly elaborate written composition, and except for a certain ‘allowance’ for embellishments and improvisatory elements, there isn´t really much wiggle-room where melody and harmony is concerned. It is not really part of the performer´s domain.

Tempo and dynamics

Thus, the relative and less fixed elements of a 19th century composition are the dynamics and the tempo with all their facets. This is where our notational system and consequently the communication of the composer´s ideas are the least advanced. This is where our rich performance tradition displays such a wide variety of interpretation, and where the composers usually expected the performers to communicate a personal narrative of their own. This is where we, the performers, come alive, where we explore the terrain, where the relative nature of these elements open never ending possibilities of interpretation, translating the written code into a personal musical testimony. This is where we must contribute as performers to communicate a persuasive narrative and a gestural content.

Of the two, (dynamics and tempo), I believe most modern performers would straight away consider dynamics, including sound production and internal voicing, as perhaps the most important parameter in the shaping of a musical idea or statement, and in separating important ideas from less important ones. Dynamics is the most instantly recognisable indication of some form of hierarchy of musical elements, both in terms of general dynamics and relative internal balance. Playing something louder than something else is a very simple way of making something appear more important. Louder in the context would mean louder than the previous bit, or it could be louder than something else sounding simultaneously, in a chord for example. We could of course even look at it the other way around, and direct the listener´s attention just as effectively by pulling back dynamically.

Less instantly understandable is probably the very sophisticated use of tempo modification we see with so many great performers, and the importance of this element in formulating a musical narrative. It is the single element of music performance which no doubt has changed the most when comparing a performance from the early 20th century against a mainstream performance of the 21st century. Performers of the time were of course as different from each other as we are today. However, on a quite different level than today, it was normal procedure for performers to introduce their own tempo modifications in a musical narrative, to separate the important from the less important in their interpretation, and to create a musical relief through internal tempo relations.

This can be found on a large-scale level and on a more localised level. We unfortunately don´t have any recordings of Mahler conducting (only piano rolls), but he was reported to have conducted several of his themes slower than most other conductors. Still, the overall durations of his performances were evidently shorter than usual. The reason, according to Mahler himself, was his lightly passing over the less significant material (More on the subject of tempo modulation in ‘Tempo and Tempo modulation’).

Creating musical relief through tempo modification is also something which appears on a very localised level. This is perhaps even more important to the realisation of a musical narrative. The instant dwelling on, and passing lightly over the more and less important material is a very typical feature of Grieg´s and his contemporaries´ performances. Butterfly is a brilliant example of this:

This temporal distinction makes the ear and mind focus on the overall line in a quite remarkable way, and creates a coherence in this section, which we have never heard since.

We certainly find this performance element in modern performances as well, and the principle is well known to most musicians. In modern performances there has however been a smoothing out which blures the musical relief, compared to for example Grieg´s performances. We are more often introduced to performances where the temporal distinctions between various elements are disappearing, and consequently we risk missing the important structural points, and important points of characterisation. When everything becomes equally important, as often is the case, nothing really stands out anymore. Again, this is not unknown to musicians of today, but we see a far greater level of sophistication in this respect when returning to the performances of the early 20th century.

Our possibility of hearing composers such as Grieg, Rachmaninoff, and many others performing their own music is of course of particular interest. It gives us the possibility of observing their level of fidelity to their own scores, and it gives us an understanding of what their concept of ‘performer-added’ tempo modification was like.

Glenn Gould is an example of a ‘meta-performer’ exploring the ‘negative’ of an image, searching outside of the stylistic footprints appearing in the written and the aural sources. Gould does not relate to the ‘marked’ and ‘unmarked’ (cf. ‘Approaching a Performance Style’) of a style in the way we usually approach the historic repertoire, and there is often less emphasis on the narrative content of the music. Gould often brilliantly illuminates other aspects of the music than we expect, sometimes going straight against, rather than interacting with a historic performance tradition, and thus playing with the expected in a different way.

The results can be totally absorbing in certain parts of the repertoire, such as the polyphonic works by Bach. Other parts of the repertoire, such as Grieg´s music, clearly suffer more from his treatment of it. The music is far too dependent on the performance tradition in which it was created. This is Gould´s version of Alla Menuetto from the Piano Sonata, the furthest away from Grieg that we can possibly get:


In his book ‘Über das Dirigieren’, Richard Wagner gives his highly articulate ideas on tempo and music-making:

It is essential that the style of execution shall agree with the matter set forth – that the tempo shall be imbued with life as delicate as the life of the thematic tissue. We may consider it established that in classical music written in the later style, modification of Tempo is a sine qua non.” (1869)

Wagner is not discussing the level of fidelity to the composer´s indications here, but the performer´s further treatment of the ‘raw material’, about performer-added tempo modifications. There were certainly various schools and great individual differences between performers in this respect, and Wagner was indeed a representative of a view not shared by everyone. But when hearing Carl Friedberg (a student of Clara Schumann, who in this respect was Wagner´s counterpart) playing Brahms, the difference to a Liszt student like d´Albert is not what appears the most striking to us (Liszt and Wagner were very much in line where tempo modification is concerned). With our 2011 ears, the really striking experience is hearing Friedberg´s Brahms and for example d´Albert´s Beethoven, and compare these performances with today´s performances of the same repertoire.

Wagner´s words about tempo modification are of course words of great satisfaction to some of us, but if we for a moment only find ourselves confident of picking tempi at liberty, we are quickly put in place by Wagner: “nothing can be more detrimental to a piece of music than arbitrary nuances of tempo, etc., such as are likely to be introduced by this or that self-willed and conceited time-beater, for the sake of what he may deem ‘effective’ ”.


After listening to such a large number of early 20th century performances, we do see another very generalised tendency, no doubt with many exceptions. When listening to many of these greatest performers of the century, we find a constant tendency of pushing beyond the ‘comfort zone’ in their music making. This is often characterised as less orderly playing, which certainly is the case at times. There are for example recorded performances by the great Alfred Cortot or indeed the later Wilhelm Kempff which probably wouldn´t take them beyond the first round of any international piano competition today. Let us side track for a moment for a charming exploration of the seventh movement of Schumann´s Kreisleriana:

The reality though, is that there are a vast number of early recorded performances with a level of accuracy that equals anything we hear on the concert stage today. But, we find the same performers constantly pushing their own boundaries, with regards to gesture, expression, or tempo. We can hear many of the greatest performers of recorded history, on the top of their careers, still pushing their playing to the edge and beyond – making music with the strongest of expression, playing as if life itself was at stake. It is an artistic attitude that affects all aspects of the music making, but the strongest effect of it is probably in the treatment of tempo.

Listen to Alfred Cortot in Chopin Piano Sonata B-flat minor, 1st movement, second theme:

There is little doubt that the recording industry has had a role in the development of performance styles over the last decades. This is partly because of the great influence specific renditions have on students and musicians around the world, and partly because of the promotion of certain performance ideals. These are ideals of order and accuracy, to some extent created by the increasing demands on the recording as a commercial product.

We have gradually seen new performance standards developing according to technological achievements in the recording industry, and it is in my view a development which is traceable even in live performances. There is also no question that these standards, and the common consequent priorities in a recording and editing process, gradually take the recording further and further away from the live performance. It is extremely important to be aware of these mechanisms in the process of recording and producing, and try to use the technology to our advantage without loosing track of the performance values we are discussing here.

Our great interest in the recordings of the early and relatively early recording period stems from this striking feeling of immediacy in so many performances. There is a freshness, a boldness, and a confidence of being part of an alive performance tradition. The repertoire we find in these recordings is often contemporary or from the near past, and in this repertoire specifically, the performers are free of the bonds of a ‘true performance tradition’ developed or rather ‘semented’ over centuries of performance and decades of recording.

When did we last hear a ‘de-cultivated’ recklessness like Mengelberg´s and the Concertgebouw Orchestra?:

Mengelberg´s recklessness is indeed very remote to the carefully measured and cultivated Brahms we hear in this reference example:

Even though we admittedly have a crush on early recordings, there is no reason to dismiss our greatest performers of today. When did we for example ever hear the line followed through the cadenza of Rachmaninoff 3. Piano concerto as ruthlessly as Martha Argerich?:

What we hear in these performances is far beyond professionalism. We hear expressive archetypes and structural ideas bypassing any paper structure, safety considerations, technical obstacles, commercial considerations, or collegial considerations regarding right or wrong. There is a de-cultivated immediacy, paradoxical in a way, because it most likely is the result of an endless cultivation. This immediacy in the gestures, the musical shaping, and the expression, is striking. And it causes a physical reaction in us.

In the best of ballet, we see this physical manifestation of musical gestures. The physical gestures support the musical gestures, and help communicate the gestural content of the music. Dance can even help realising a gestural content, which is not so strongly apparent in the musical performance itself. In a musical performance not supported by this visualisation, or interpretation of musical gestures, the musical gestures are even more important to realise. Music on its own appears poor and irrelevant to us if the gestural content isn´t strongly apparent, and there is a heavy responsibility on the performer in realising music´s gestural content.

An interesting and highly relevant side track to this is the recent finds on the mirror neuron system of the brain, and its role in our ability to recognise the meaning of, and the intention behind a communicative signal. Our neurological response to language, music, and physical movement seems to have very much in common, and is activated by the mirror neurons. I suppose we have all experienced music´s physical and emotional power. The most interesting part is what happens when listening to music: “the human mirror neuron system enables us to not only recognise, but activate the same physical actions and emotional reactions.” (Istvan Molnar-Szakacs and Katie Overy: Music and mirror neurons: from motion to ’e’motion) This explains to some extent why the recreation process is valuable, and it explains what we already know: music is capable of evoking emotional and truly physical reactions through its gestural content. This is not ‘New Age’!

Our recreation project has given a very special opportunity to experience this point first hand. Grieg´s recordings were made late in his life, and his health was deteriorating. Still, keeping track with him proved to be an extraordinarily difficult task. His tempi are always brisk, and keeping track with his often reckless gestures, sudden moves, and strongly shaped and insistently pushed phrasing, turned out to be a never-ending battle. ‘On the heels of Grieg’ could very well have been an alternative title of our project.


When considering the narrative qualities of Grieg´ performances, the vocal qualities stand out as particularly important. In trying to capture Grieg´s particular style of playing the piano, we gradually realised that the vocal qualities of his playing were present practically all the time. By constantly referring to Grieg´s performance, we always found the spoken qualities more present in his than in my own performances. There was always the extra mile to go. This is perhaps not so surprising to find in a piece like Remembrances or the Alla Menuetto. What really fascinated and surprised us was his ability to keep the vocal or spoken qualities even in the fastest pieces like the example above. This is a very subtle aspect of his performances, which may not be immediatly apparent, but nevertheless very significant to his style. Capturing the spoken qualities in for example the first theme in the example above proved extremely difficult.

A really lovely example of this is found in Butterfly. Grieg´s performance appears to be very rhapsodic, and his version is a pretty fast one. In the most beautiful way, Grieg still manages to keep the spoken or vocal qualities throughout the performance, even in the rapid semiquavers. This really distinguishes Grieg´s performance from any other:

The spoken qualitiy here is not just a matter of a singing, basic piano sound. It is also due to the beautiful ebb and flow of the line which so much resembles the natural breathing. Even when Grieg is placing a passage in the shadow of something even more important, as in Butterfly, there is a spoken quality to the smallest details.

As we heard in Ferrucio Busoni´s Chopin-performance (‘Ambiguity and Multi-layeredness’), the vocal qualities of his piano playing are striking, too. In his suggestive advice to pianists (quoted by Henry T. Finck) we hear the resonance of Schumann:

The bar-line is only for the eye. In playing, as in reading a poem… you must speak the piano.”