It is not my intention here to go into a thorough discussion of Lendvai’s theories, but just to point out that there are constructional principles with which the performer should be acquainted, and upon which he/she should reflect critically. Whether this will influence the playing, is another issue, to which I shall return.

Thematic and tonal structure

More immediately relevant when it comes to the performer’s knowledge, is how the thematic material is transformed as the piece proceeds. Fig. 3 shows the metamorphosis of the first theme of the Allegro, whose genesis is found at the end of the introduction (bar 20) and which occurs in the two other movements as well. We can therefore speak of a cyclic theme.

The theme evolves from three repeated notes in bar 20 via 4 notes to the 5-note motive of bar 26, which finally becomes the first theme of the Allegro in bar 32, with 6 values, three crotchets and three quavers. There is also a form consisting of two crotchets and three quavers, and the diagram shows the metamorphosis of both into the form of bar 217. Here, the initial quaver rest is replaced by a note, which can be seen as a permutation of the original durations. Yet another form is encountered in bar 386, expressed as an alternation of quavers and crotchets. 

In the second movement, the six original attacks have become of equal length in the form of a quintuplet with final note. Of the original melodic shape, only the minor third is kept, now upwards rather than downwards. In the third movement, it has become a purely rhythmical theme played by the drums (6 attacks).

[4] Anssi Karttunen: How did it originally sound? [back]

[5] Daniel Gregory Mason: The chamber music of Brahms. Books for Library Press, Freeport, N.Y., 1933. Reprint 1970, p. 43. [back]

[6] Brahms the progressive, in Style and Idea, Philosophical Library, New York, 1950 [back]

[7] Walter Niemann is one of them. - die ganze elegische zweite (f-moll) und geheimnisvoll gedämpfte dritte Themengruppe (cis-moll) - Niemann: Brahms. Schuster & Loeffler, Berlin 1920, page 242. [back]

[8] Cited in Dag Østerberg: Brahms. Gyldendal, Oslo 2003, page 116. [back]

[9] See the rehearsal video at 23’48. [back]

[10] Rehearsal video, 02’44 [back]

Does thinking enhance performance?

Account of a process leading up to the performance of Bartok’s Sonata for two pianos and percussion, and Brahms’ Piano Quintet. 

- By Håkon Austbø

The history

The question whether thinking, and in particular analytical thinking, can make musical performances more genuine and unique, is the core matter of The Reflective Musician. To be sure, we are not so much concerned about performances being better, but simply about whether this knowledge does anything to the performer, and therefore to the performance.

To test this, we have set up several processes during the two years of the project. One of them was carried out in the spring of 2014. We set out in the autumn of 2013 to find a group of musicians to rehearse and perform Bartók’s Sonata for two pianos and percussion and Brahms’ Piano Quintet in f minor. These pieces were chosen because both composers share a strong intellectual involvement, combined with extraordinary expressive power. The compelling constructive logic of both pieces enhances this expression rather than standing in its way. Therefore it seemed fruitful to investigate whether insight into these constructive principles would enhance the performance of the pieces.

The two groups of musicians had one common performer: myself. Then project group member Ellen Ugelvik was a natural choice to play the other piano part in Bartók, since we had worked together in different settings already. Hans-Kristian Kjos Sørensen was invited to take one percussion part since he already had played it with me, and Kjell Tore Innervik was asked to join him. For the strings, some of the participants changed during the course of events, and finally the group consisted of:

Terje Tønnesen and Ragnhild Hemsing, violins

Jon Sønstebø, viola and Audun Sandvik, cello.

These are all first class performers who were willing to go into the special process required here, different from the normal procedure of rehearsing and performing a piece. The idea was to confront the performers with some of the analytical thinking and to generate discussion and reflection through doing so. We also intended to make the rehearsals accessible to an audience and to document them on video. This would open up the process, make the participants more vulnerable and thus sharpen the performers’ awareness. All this was made clear to the participants from the beginning, and they all accepted.

The open rehearsals were held at the Norwegian Academy of Music during the month of April, 2014, and the performance took place in Sandvika near Oslo on April 30. All these events were recorded on video, and the edited recordings are accessible on this page.

The Bartók Sonata

Bartók the intellectual

I first analysed Bartók’s Sonata for two pianos and percussion (1938) in Paris, where we had installed an analysis class following the May ’68 évènements. We studied both this piece and the composer’s preceding masterpiece, the Music for strings, percussion and celesta, and I was so fascinated by Bartók’s compositional mastery that I chose the latter for the following year’s exam. This pair of works marks the culmination of Bartók’s integration of intellectual construction, folkloristic elements and basal primary forces. He became for me the prototype of the reflective musician, reaching a synthesis of intellectual and spiritual force achieved by the very great only. It seemed that after having reached this high point of his oeuvre, and then fled Nazism and immigrated to the United States, his creative force dropped ever so slightly, although works like Concerto for Orchestra or Piano Concerto no. 3 are among his most popular works. I think there is reason to state that he never again achieved the sublime mastery shown in the two works from 1937 and -38.

In the Paris analysis class we studied the theories of Ernö Lendvai, then quite new. His standard book was yet to be published [1], but his theories had started to spread among musicologists. Lendvai sought to prove that much of Bartók’s creative force was centred on two things: the system of axis and the golden section. I will assume in the following that the reader is familiar with these theories, and otherwise refer to Lendvai’s book.

The golden section and Fibonacci, according to Lendvai

Today, it seems that Lendvai carried his conclusions somewhat too far, but the core of his analysis is difficult to ignore. When he refers to the first movement of the Music for strings as entirely based on the series of Fibonacci (where each term is the sum of the two previous ones), it is fascinating. However, it disturbs the joy of the analytical mind to discover that the addition of an empty bar in order to reach the number of 89 bars is necessary to prove his point. True, the climax and the following inversion of the theme come in bar 56, and this is very close to coinciding with the Golden Section (GS), but it doesn’t strengthen the theory when facts need to be twisted to fit it.

It is also true that the total of quavers in the Sonata for two pianos and percussion, divided according to the GS, coincides with the transition between the 1st and 2nd movement. Whether it falls after the last note of the first movement or on the first note of the second movement, however, depends on how you count [2]. Also, the GS of each movement comes very close to an important turn in the form. All this was indeed fascinating to discover, and I still think there is reason to claim that Bartók was seeking a universal constructive principle of proportions. It is therefore one of the elements that I took into the discussions at the open rehearsals.

When it comes to the introduction to the first movement, Lendvai again shows the division according to GS, juxtaposing “positive”(long) and “negative”(short) sections. Some of it fits into Fibonacci numbers, but not all. The quaver count until bar 18 (Un poco più mosso) looks like this:

Although the first two sections here fit with the Fibonacci series, the third doesn’t, and this brings the total count of the large sections to 85 of 138 rather than 89 of 144 (the numbers we discussed in Paris). What remains, however, is that the relationships between the various short and long sections correspond very closely to the golden section.

The use of Fibonacci numbers is also evident when it comes to intervals. That the typical Bartókian chord C-Eb-F#-Ab-B consists of the constellation 3+5=8 placed symmetrically, is clear (See fig.2). The main theme of the first movement Allegro responds to such numbers in the ambitus of the four phrases. The two phrases of the second theme extend the ambitus further to 21. The chord at bar 99 reflects the same stretch. All this is seen in fig. 2.

In all these transformations, the rhythmical aspect is by far the most important, and those rhythms must be understood and felt by the performers. There are other thematic connections of importance. The third theme of the first movement is based on the same major sixth as the third theme of the third movement, and the second theme of the first movement use minor sixths both vertically and horizontally:

Note that the third theme in the recapitulation takes on the form of a minor 6th (D-Bb instead of D-B). Several themes are also combined at various places. The most striking is perhaps this passage in the development of the first movement:

The first theme is here superimposed upon an ostinato built on the introduction theme. Then, in combination with the rhythm of the first theme, we get a series of accents forming the melodic line of the second theme. Although more hidden, the rhythm of the third theme is also heard at this point.

Another aspect of importance to the interpretative process is the tonal structure. Here again, Lendvai has formulated the principles. Much of Bartók’s harmonic language is based on the contradiction between fifth and tritone, which can be illustrated by the circle of fifths where the opposite poles form the tritone axis (C-F#, as in this work). The poles situated at 90 degrees from these form the secondary axis of the tonic (A-Eb), and the remaining eight notes of the chromatic scale belong either to the subdominant or to the dominant axis. This way we get an organization of the twelve tones into a tonal system, as opposed to Schoenberg’s non-tonal serial procedure. The first movement goes from an opening F# to C during the introduction. The second theme is centred on E in the exposition, that belongs to the dominant axis and is also the centre at the passage shown in fig. 5; the main section of the development. In the recapitulation, the second theme comes back in A, responding to the classical scheme: back on the tonic axis.

In the second movement of the Sonata, the main material is centred on B-F and the middle section on D-G#, which gives this movement a clear subdominant character. The third movement is based on diatonic harmony and less on the axis.

This last movement uses the “acoustical scale” that stands in clear opposition to the axis system, although the tritone forms a common point between the two. As Lendvai points out [3], the theme here represents a straight line whereas the theme of the introduction has a circular form:


This is but one of several dichotomies present in the piece. In the video from the open rehearsals it can be seen how our “opponent” Emil Bernhard brings this into the discussion. The circular- versus-linear movement can be interpreted as representing darkness and light and thus reflecting important psychological factors in the work. The chromaticism of the axis-related material emerges as gloomy, dark and dramatic, while the acoustical scale or other diatonic material represent the optimism, the light, the joy, often related to folklore. To this latter family of materials also belongs the second theme of the first movement, with its ‘Bulgarian’ rhythm (see fig. 4).

There is another obvious dichotomy present in the very instrumentation of the piece. Although the piano can be called a percussion instrument and is sometimes treated as one, it is much more a melodic and harmonic one. Inside the percussion array, there are also melodic instruments, the xylophone being the most prominent here (and, to a lesser extent, the timpani). Still, there is a clear dichotomy between percussive and melodic elements, most evidently presented in the beginning of the second movement, where the uncompromising sound of the drums contrasts with the imploring yet distant cantilena of the pianos. This is another circular theme, and we are again caught in a stifling, claustrophobic atmosphere:

For the performers, the obvious challenge here is to marry the melodic character of the pianos with the mechanical one of the percussion. This has to be dealt with while rehearsing, and however small a degree of phrasing is used, the percussion will have to accommodate. At other points in the score, the sonorities of pianos and percussion are, on the contrary, fused into novel sonorities. Four instances of this are shown in fig. 8.

In three of the examples, the percussion instruments, xylophone, side drum and triangle respectively, enhance the sharpness of the piano sonority, whereas in the second example, the tam-tam and pianos enrich each other’s harmonics, which results in the colourful, vibrating sound of this Fibonacci-based chord. Such sonorities can only be achieved with this instrumentation and must be optimized during rehearsals.


Influence on playing

Awareness of such facts is more physical than that of abstract or constructional ones, and therefore more directly applicable by the performers. But also, when it comes to the knowledge of crucial points in the form determined by the golden section, it will help the performer to give these points their full importance and convey their dramatic impact. In fig. 9, these points are shown for the three movements. 

In the first movement, the GS falls on the recapitulation of the first Allegro theme, or the cyclic theme, in bar 274. This version of the theme is an extension of the first phrase into four bars, where each of the first three attacks is repeated and extended to the total duration of a semibreve. Melodically, there is an extension of two chords at the end, D-Eb, so that the theme strictly speaking ends with the C just before this, then describes an aborted inversion. (See also fig. 3.) For the non-reflective musician, the theme might not be recognisable here.

In the second movement, the GS point at bar 57 falls inside a passage that builds up to a climactic point five bars later, where an element from the first movement is brought in. Why didn’t Bartók rather put the golden section here? Perhaps, the whole passage is to be considered as a turning point and the division is therefore kept approximate. A striking fact of this movement, by the way, is that each of the three sections in the ABA form have exactly the same chronometrical length, 1’53” (the last A is 1’52”, close enough). Bartók indicates the precise duration of each section of the work. This underlines even more dramatically his meticulous sense of proportions, and prompts the performers to keep the tempi very exact.

In the third movement, the point of the GS is, again, the recapitulation. Here also, the theme is not immediately recognizable, since it is inverted and fragmented as well as treated in canon. It comes after a tension-building section where the opening interval of the first theme is repeatedly stretched out against an ostinato of chromatically ascending semitones. Intuitively, one is led to bring out the descending fourth in the xylophone and second piano as a dramatic climax, and this can only be strengthened by the awareness of an important structural point.

[1] Ernö Lendvai: Béla Bartók: An Analysis of His Music. Kahn & Averill, London 1971 [back]

[2] The count is 6432x0,618=3975. But in my count the section falls after the last note of the first movement, whereas for unknown reasons, Lendvai places it on the first note of the second. The first movement comprises 3481 quavers. Lendvai, p. 26. [back]

[3] Lendvai, p. 77. [back]

The Brahms Quintet

Complexity and resistance

The Brahms Piano Quintet in f minor, op. 34 was first conceived as a string quintet – later destroyed by the composer – then as a sonata for two pianos, published as op. 34b, before he decided to rewrite it for piano and string quartet. It seems that it didn’t gain immediate popularity, although this appears strange today when it is often acclaimed as one of Brahms’ greatest masterpieces. Perhaps the fact that he struggled to find the ultimate intrumentation for the work, shows that it is a piece of music not so much bound to a specific sonority. In fact, it works even well in the two-piano version, that I have also performed many times. Anssi Karttunen has recently made a reconstruction of the original setting for string quintet with two celli (as in the Schubert C Major quintet), and states:

Neither of the existing versions is quite as inventive in their use of instruments as one would expect from an original piece by Brahms, even at 29.  This is evident if one compares the piece with the three Piano Quartets or the String Quintets with two violas, which could never have been written for another ensemble. [4]


Is the less idiomatic writing in this work of Brahms, compared to the other works, a reason for its late recognition as a masterpiece? I think it also has to do with the way in which he treats the material. Although already in his very first works, the piano sonatas, he shows a strong concern for intellectual, constructive principles, the way the motivic work permeates every aspect of the creation evolves gradually, and by the time he composed the Quintet, his mastery in this craft was absolute. With this mastery, he crafted a very complex piece, even with today’s ears and this could be one of the reasons that some people have found it less accessible than some of his other chamber music works. D.G. Mason writes about this:

The great Quintet was one of the slowest of all Brahms’s works to win recognition; undoubtedly it is one of the hardest of them all to understand. [5]


The author attributes this difficulty to the profusion of themes, especially in the first movement. But the mastery of Brahms lies precisely in his ability to draw material for such a multitude of themes from the same basic forms, often presented in the very first bars of the work. True, he had attempted this all his life, and early works such as the three piano sonatas and the Schumann variations op. 9 already contain complex motivic and contrapuntal work, but this sometimes feels as a result of wanted intellectual exercise, rather than springing from a natural source.

It is interesting, for the understanding of this development, to compare the two versions of the Piano trio op. 8. When Brahms revised it in 1891, he must have felt that all the excellent material warranted countering by a richer array of complementary elements. In fact, he kept the first themes of all movements and almost all of the scherzo (except the ending). When one studies the contrasting themes he composed for the second version, it is striking that they are just that: contrasting, whereas the original themes offered little variation (except, maybe, an Allegro passage in the slow movement that was rejected) and bore witness of great resistance. The resulting revised version emerges on the contrary as an incredibly rich, natural and balanced whole. This was how far Brahms had come thirty years after conceiving the Piano quintet. (See also our discussion of the piano pieces op. 118. LINK)

Thematic connections: Grundgestalt

It becomes mandatory, then, for an understanding of the Piano Quintet, to find out how the themes relate to each other. The use of Grundgestalt that Brahms came to make one of his compositional trademarks, was, as with many of his aesthetic principles, derived from Beethoven. Schoenberg continued along the same path, and formulated a terminology of such techniques in his didactic writings. He also praised Brahms as a progressive composer [6] against the ruling opinion of the time that he was a conservative while Wagner was the progressive.

Most of the material for the Piano Quintet is indeed found in the first four bars of the first movement. In order to see through the complexity of the thematic material, I have made a survey of the principal themes and other material that may be found in the image on the right. I shall try to elucidate the diagram in the following, highlighting some details at a time since the total survey may appear confusing.

To avoid misunderstanding, I don’t adopt the labelling of some analysts of the theme at bar 23 of the first movement as a second theme [7]. The convention when it comes to sonata form is that the second theme is in a contrasting tonality and returns on the tonic in the recapitulation. The theme at bar 23 starts in F minor and must be called bridge, whereas the true second theme comes at bar 35. It is in C# minor in the exposition and in F# minor in the recapitulation, then finally restated in f minor in bar 208. I will discuss these tonalities further on.

The central motif, designated ‘a’ (top left) can be seen as a combination of two elements (elementary particles, as it were): the ascending fourth C-F (q) and the three consecutive notes F-G-Ab (in blue). Both will be treated together and as separate elements. They will generate a portion of the second theme at bar 36 in the first movement, which will in turn generate the first theme of the third movement. Then, they will form the basis of the third theme in the first movement (bar 74), treated in canon between the two violins. The motif is also the origin of the theme of the trio of the third movement (bar 194) and, finally, of the main theme of the fourth movement at bar 42.

Noteworthy is that the second statement of a, here called a’, omits the F and therefore starts with a fifth rather than with a fourth. This will give birth, in its turn, to a number of events in the piece, for example the one shown here at bar 53, where the fifth is the result of a developing stretch in the two preceding bars. At bar 260 of the first movement, it is not an inversion of the theme, but a transformation of its second bar where the ascending fifth is turned into a descending fourth. Hence, it would be wrong to phrase the first bar into the first beat of the second, since this would conceal this motivic structure.

Nevertheless, one still meets string players who think the second version is superior because it allows for a better bowing.

It is the three-note motif that, in retrograde motion, ends the quintet. In this form, the a motif was already used for the bridge theme of the last movement:

This use of the two most important cells of the piece, a and c, to sum up the piece at the end, gives a strong feeling of unity. The fourth C-F that also belongs to the initial form, is heard in the bass for the final cadence. In the main theme of the first movement, F-C are the second voice.

Structural motif

It is time to discuss the use of the D flat, shown in the survey as motif c (in red), as it tends toward its resolution on the C a semitone below. Fig. 4 shows the origin and the main instances of this motif through the work, together with the octave motif d.

The motif c is not just a melodic one, it is also and above all harmonic and structural. It represents the minor submediant and its conflict with the dominant. As a tonality, it yields in its minor form the key of the second theme of the first movement, as it appears in the exposition at bar 35, enharmonically notated. In the introductory bar, the accompaniment is heard, containing the motive c in the new key. This gives a combination of the major mediant and the minor submediant of F, and results in the augmented triad (F-A-C#) so often used by Beethoven in his choice of keys. The middle part of the second movement is in E Major, which stands in a similar relation to the main key of A flat.

The D flat plays a role at many structural moments in the Quintet, of which I have only pointed out the most important ones in Fig. 4. While it is not very prominent in the first bar, it becomes the focal point of the drama already in bar 5 and brings the first theme to an end in bar 22. The same form is immediately reiterated in the next bar to commence the bridge theme (called second theme by some analysts, see above), lending its soaring character from precisely the conflict between Db and C. The tonality of D flat concludes the exposition after having dominated 57 of its 90 bars, either in the form of C# minor (second theme) or as D flat major (third theme). At the end of the development, where the motif c, as part of the second theme, has played a dominant role, the D flat returns in the cello and is taken over insistently by the piano to bring back the dramatic surge of bar 5:

In the Scherzo, the motif is found in two of the three themes as a dissonant neighbour to the dominant – the form c2 occurs in the second voice – and since we are in c minor, it is an A flat (becoming A natural in the third theme bar 23 but then losing its dissonant character). However, in the course of the scherzo, the music will tend toward F minor, and as the first theme returns in this key in bar 178, the conflict between Db and C becomes acute and culminates in the form seen in bar 193. That we are on the dominant of F minor, not in C major, becomes clear only with the transition into the fourth movement. The trio continues in C major and doesn’t resolve the conflict.

This tonal instability is important for the musicians to feel and was thoroughly discussed in the rehearsals. At the same time, this last figure with semi-quaver and dotted quaver is a tribute to Schubert’s String Quintet with two celli in C major, that may have been a model for the first version of this work, scored for the same ensemble. The end of the Scherzo here is a direct quote from the Scherzo of Schubert’s quintet. This may be a reason that Brahms puts such energy into this ending of the third movement.

Then he goes on with one of the grandest passages he ever conceived. The introduction to the fourth movement takes the semitone tension into epic dimensions. It starts with the octave F-F in the cello (the resolution of the dominant tension on C), followed by Gb, a new way of combining the octave jump d and the semitone c. This combination was seen in the second theme of the first movement, and the octave that initiates the middle part of the second movement is extended to a minor ninth in bar 56, leading to the semitone circling-in of D in bar 66, which is the dominant of the wrong key, discretely shifted to the real dominant E flat in bar 71. This was a striking use of the c motif as a tonally structuring element, but when it returns in the fourth, it will take us through the most remote tonal regions, until the clash in bar 13; the Db of the first violin on top of a pulsating dominant 9th chord on C in the piano, extending into a melodic passage of great intensity:

This melodic theme has not been heard before, nor is it to be heard later in the piece. It does relate to other themes, though. There is an almost exact retrograde of its interval structure in the second movement:

After a shortening of the phrases from 4 to 2 bars, then to 1 bar, the liquidation process ends with a violent E minor chord (a return to the key of bar 12) alternating with Bb minor. The Db bass of the latter then descends into C in bar 35 to prepare the theme of the Allegro non troppo, while the principal theme of the movement is gradually taking shape.

After this, the D flat is not heard for a while, except for a short passage in bar 211, until it returns in the form of C sharp minor in the final Presto, non troppo. The C sharp minor played an important role in the first movement, here it colours a variation of the principal theme, before the enharmonic Db descends into C for the stormy reiteration of the variant in f minor.

Surprising parallels

There is still one principal motif to be discussed, which I call b. It appears in the first bar overlapping with the a motif, and in its primary form it extends to the Db, becoming the pivot between the stepwise motion of a and the arpeggiated figure that will be extended during the following bars. The main stages of its history are shown in Fig. 9.

Many of the themes in the Quintet are generated wholly or in part by this motif, either in its original form or in its rudimentary form b1 and derivatives. This form, consisting of a rising second and a falling third, is as much used by Brahms in his whole oeuvre as the three-note motive a, and is in fact nothing else than a permutation of those three notes. (See the discussion of op. 118 elsewhere in this exposition.) The arpeggio form returns immediately in bar 5, already discussed, then in the end of the first movement in a sequential, downward motion (bar 290), then again in the middle section of the second movement (bar 34), slightly modified but with the same extensions of the intervals as in the opening theme.

In its rudimentary form, but retrograded (b2) it forms the principal theme of the second movement. Pitchwise, it is repeated twice, and rhythmically the durations are shifted to the left in relation to the beat, with the second one shortened; an interesting rhythmical procedure for Brahms. In the second phrase of the theme (bar 23) the motif starts on the upbeat and is in even quavers. This form comes back to constitute a second section of the middle part at bar 47.

We also see from fig. 9 that the form b2 is a constitutive element of all three themes that follow each other during the whole scherzo, here marked as A, B, and C. The motif, together with the alternate form of c, is in fact the element that bind these themes together. It can be seen, too, that theme C is a major, augmented variant of theme B.

The last, surprising fact that we see from this survey, is the obvious link between the main theme of the second movement and the second theme of the fourth (bar 94). This theme starts with the downward minor second and fourth which is a retrograde of a, called a1. Then follows three notes in a characteristic rhythm, in fact that of the theme of the second movement. Since the intervallic structure is identical, this connection is most certainly intentional and will undoubtedly have some influence on the way it is played (it is often taken too fast). The a1 motif in the end of the phrase, by juxtaposition and sequencing, then becomes a’ (downward fourth and upward second), a form that has been heard in the first movement (see fig. 1), and which is then treated in canon and diminution (bar 125), and, combined with the upward motion 2xa, forms the material of the energetic contrapuntal passage that follows, and finally crowns the triumphant restatement of the secondary theme in the elaborate coda (bar 404).

Formal questions

If I have insisted so much on the motivic work in the Quintet, it is because this is such a distinctive feature of Brahms’ conception of form. The unifying effect of motivic connections is one aspect of this, the constant transformation of the material another. In the mature works of Brahms, one idea constantly and seamlessly grows into the next and results in perpetual development. This was one of the main points of criticism by his contemporaries, as that of H. Zopff in Neue Zeitschrift für Musik:

He starts to work on his theme already before having rounded off its presentation, and in this way fails on beforehand in giving it the full power of a genuine impression.. He presents his ideas for us, sometimes two, three times in a row, and buries himself in such passionate diggings that even his brilliant contrapuntal mastery will seldom be fully appreciated. [8]

Is this not a characteristic of the mature Brahms that is already present in a work he wrote at 29? Indeed, because by then he had gained the mastery of this method, tried out and experimented in a number of earlier works (I mentioned a couple of these earlier). Today, this feature must be seen as being ahead of its time; it became more or less the ideal of the Second Viennese School and its followers.

The forms inside which he applies this processing, are the classical forms of sonata, Lied, and scherzo with trio. There are some unorthodox points in these forms, though. I already mentioned the tonal scheme of the first movement, and how the choice of C# minor for the second theme springs out of a fundamental conflict in the work, that of D flat against C. In the recapitulation, the theme is first stated in F# minor, a logical parallel to the exposition. But through a shift from C sharp to C in the bass (motif c!) a variant of the theme is heard in F minor:

This variant gives a dramatic change compared to the original: the octave jump has disappeared and the character become utterly desolate, devoid of its dancing quality. Then, after the rest of the themes have been confirmed in F minor, a passage in F Major, poco sostenuto, brings the variant of the first theme discussed above (a’ in fig. 1). This passage at bar 261 gives a moment of repose and distance before the passion breaks through and paves the way for the stormy coda.

The second movement is in fairly straightforward ABA form except, perhaps, for the transition at bar 55. It contains a development of the octave motif d in combination with the semitone of the c motif ending up in the cello passage at bar 64 shown in fig. 4. But, against that, a rhythm is heard in the piano which is the one used earlier for accompanying the first theme; this time, however, with an independent melodic content (b2) that somehow makes it a principal voice.

The scheme of the third movement is unproblematic and shown below. As I pointed out earlier, the three themes A, B and C are constantly juxtaposed:

A - B - C (C min/Maj) - A (C min) - B (G min) - Fugato on B (modulating) -

B - C - A (Eb min/Maj) - A (C min) - B - A (F min)

What is unusual here, is the tonal scheme, as I discussed above in connection with the ending on Db-C that is in fact a dominant. The trio consists of a broad melodic theme in C Major, based on the a motif (see fig. 1) and a contrapuntal development based on b2.

The finale is a rondo with both an introduction and an elaborate coda, not an unusual fact in itself, nor is the transformation of the theme shown in fig. 8; similar transformations are found in several of Mozart’s and Beethoven’s finales. Two points in the form are unusual. At bar 161, returning from the stormy passage following theme 2, we get a pseudo-restatement of the principal theme, but in C minor and varied, that in its turn leads back to the real restatement in F minor. Then comes a developing variation at bar 211 and the second statement of theme 2 in F (the first time was in C). At the end of the corresponding stormy passage we get the second variant of the main theme that precedes the C# minor coda (bar 322). Fig. 12 shows the variants of the theme at both points.

This last variant is devoid of life. As the cellist, Audun Sandvik, puts it in the rehearsal, ‘it is the saddest moment in music history’ [9]. Even without the analytical knowledge it is possible to grasp the psychology of such a passage. Music is stronger than words, and there is no point in analyzing if it is only to describe what is happening. Analysis is useful only if it does deepen the understanding, and again, the cellist made a strong point about this in the rehearsal, saying:

Performance at Bærum Kulturhus, Sandvika

- April 30th 2014 [back]

Fig. 9: The golden section in each movement of the Bartók Sonata, ed. Boosey & Hawkes (click to open)

Fig. 7: Beginning of 2nd movement of the Bartók Sonata, ed. Boosey & Hawkes (click to open)

In the video from the rehearsals, a discussion develops between the musicians about how to relate to such facts. The discussion is also transcribed in a separate text (right column).

It touches a crucial point of The reflective musician: Does this knowledge influence my playing? Does it make the performance different? The fact of posing those questions is alone proof that it does.

‘When we have this knowledge, we don’t have to do anything with it. Knowing the importance of the note is going to do something about the way you play’ [10]. The feedback from the musicians (right column) shows that it did something to them.

What the comments also show, however, is that this is not common practice among performers today. Many of them are unsure how to translate analytical knowledge into practical music making. There is certainly work to do here: even though such knowledge will influence the performance, whether or not consciously, the awareness of structural issues needs to be revived as an integral part of musicianship, like it was a century ago.

Fig. 3: Cyclic theme transformations (click to open)

Fig. 9: Material generated by motif b (click to open)

Fig. 1: Transformations motif a (click to open)

Fig. 1: Transformations motif a (click to open)

Skip to chapter "Bartók sonata"

Skip to chapter "Brahms piano quintet"

Feedback from musicians on the Bartok-Brahms process

Jon: The project was instructive with regards to the theoretical focus. The challenge for my part was to translate this into music, to create a bridge between knowledge and tones. The idea was good, and we probably succeeded partly, in spite of some shortage of time towards the end.

Terje: In my view it was instructive, but I think that all knowledge, albeit theoretical, socio-historical as well as emotional, is useful material in the working-out of interpretations. But for me, especially in Romantic music, the expression and speech of the pieces is, above all, an “emotional speech”, and all the rest becomes just one of several tools. Therefore, if you look at the project as a musician’s way to the podium, it was an unusual rehearsal time (in my opinion). But I expect the filmed product can be extremely interesting for students – and that was probably the purpose.

Hans-Kristian: As I said during the open rehearsals, the basic idea of “The reflective musician”, namely, which incentives are underlying the musician’s development of a personal interpretation of the music, is something very fundamental. All the available knowledge, the cognitive, are tools that can be used for this. What was difficult to get across this time (surely due to lack of time during rehearsals), was how this knowledge is used by the musician. That is: How do you go from A (theory, history, analysis etc) to B (a personal interpretation and performance on stage). What happens to the musician in this process? What does the musician think about reflecting? The meta-cognitive.

How can the knowledge be applied?

Speaking about music, or other arts for that sake, is like speaking ON the lines. But all good art is mainly about what is BETWEEN the lines.

We could have spent more time on this, to approach even more what students and others often wonder; How to speak personally through music.

Håkon: Common to these comments is that one doesn’t yet quite find out how to translate the knowledge in practical playing. For my part, I think this will remain partly unconscious, but that all accumulated knowledge will leave its imprint on the performance, without one being able to point out exactly what this consists of.

Ellen: I agree with the others that the theoretical-analytical aspects probably got too much space in this process. This is the part of the process that is clearly formulated, where we possess an established vocabulary. I won’t ponder further on this, since it has been well described by the others.

Bartók’s sonata is a work where everybody plays an equal part. All instruments exploit their percussive – and, somewhat unexpectedly, their lyrical qualities. I would have liked to try some of these percussion instruments to really feel the challenges and qualities on my own body: experience the percussion parts physically. I think this would yield something interesting precisely in this work, and in many other works of Bartok because most of it is physically intensive, even provoking fatigue.

DIALOGUE ON BARTOK (video at  24’39”)

Hans Kristian: The analysis of the whole piece is a very cognitive thing, because this is knowledge, right? We as musicians have to play this piece, and to make all this cognitive knowledge into – not only inspiration – but as tools of how to play the piece, this transition from the theory and cognitive stuff over to the meta-cognitive dimension is the trick for us. And, that’s also, I think, the trick for most musicians playing this piece, because, yes, we know this golden section and all this now, so then, how to use it, how to relate to all this huge amount of knowledge, when you just have to play a snare drum or a xylophone, you know, this is the transition.

Håkon: It is a matter of trying to get into the process that Bartok was in when he composed it. Why did he think of all these mathematical things, and also the tonal scheme, of course… it is very thought out - the whole axis system.. So, is that only an intellectual exercise, or is it a necessity to compose the piece? - and I think that’s what we have to..

Ellen: What do you want the audience to experience? Do you want them to experience this very thought-out scheme? Because it is such an emotional piece, in a way.

Hans Kristian: It is!

Ellen: - so it doesn’t really come through, I think. When you play, it seems like it’s not this type of work going on, because you are sort of above that. Is it possible for you to explain how it works, because every time I ask you: and then what? How do you play this, how do you touch the keys?

Hans Kristian: I think this is actually a major part of the thinking of The reflective musician.


Håkon: Well, we don’t have to think about it. I think, unconsciously also, when we know this, we will play it better. - I think.