Handling complex language

In this section, we consider the aspects melodic complexity, harmonic complexity, and complex sonorities.

Melodic complexity

For much of the repertoire, I found that melodies – and parts of melodies – could be applied in improvisation using my usual ways of working. Such approaches could include playing melodies with improvised variations, improvising over melodies while singing at the same time, and – as mentioned earlier – trying to learn melodies by ear and transposing them into new keys. However, some of the melodies presented particular challenges, in pieces like Variazoni, Leggiero, Sonatine for Flute and Piano, the second theme from Forest Musicians and parts of All in Twilight. This could relate to aspects such as tonal ambiguity as well as other properties; that they contained large intervals, an extensive range, and rapid shifts between different rhythmic subdivisions. In such cases, we based the improvisations on open approaches – e.g. ‘free improvisation’ – rather than adhering to specific melodic or harmonic departure points, which gave more room for improvisers to freely associate on the material at hand. For instance, when working together with en en en on Variazoni, a piece written using twelve-tone techniques, we went into an open, collective improvisation after repeating the final chord a few times. Similarly, with Leggiero, after trying improvised sections that were more controlled, we ultimately went for an open group improvisation with a predetermined ending.

Interestingly, despite – or maybe because of – the high degree of melodic complexity in these pieces, they seemed to offer a lot in terms of impulses for collective improvisation. Variazioni, for instance, often managed to ignite improvisational creativity in the sessions and live performances together with en en en. This may relate to the piece's distinctive character, where wide-ranging atonal lines are combined with rhythms that are relatively accessible, often based on eight notes. Another aspect is the contrast between score adherence and free playing; after closely following the written score for an extended time period, it is somewhat expected that improvising musicians will take advantage of the free sections as they eventually appear. 

Harmonic complexity

Some of the harmonic progressions called for certain harmonic strategies, as described below. These were: opening up the harmony, working with line-based approaches, extrapolating material from the melody, internalizing the sounds, finding common notes, and creating balance between dense and open passages.

Opening up the harmony

A common way of handling progressions that were harmonically dense was to find a more open interpretation of the harmony. This could arise out of a feeling of being locked in by the written-out harmonies, as expressed by Söderqvist when discussing our duo repertoire:

SS: I can feel that with the written-out harmony I have to struggle quite a lot. Because in a way it becomes... It restricts you, in a way. At the same time, it gives you control too, of course. ... But the risk when I play, if I'm going to improvise over the harmony, particularly more difficult and complex harmony, is that I paint myself into a corner and think that I have to include things. And then you might forget about the melodic part. (Listening session with Söderqvist, 2022-10-07)
[SS: Jag kan ju känna att med den utskrivna harmoniken så får jag kämpa ganska mycket. För att på ett sätt så blir det... Det stänger in, på något sätt. Samtidigt så ger det ju kontroll också förstås. … Men risken när jag spelar, om jag ska improvisera över harmoni, och framför allt svårare och mer komplex harmonik, är att jag liksom målar in mig i ett hörn och tänker att jag måste 'få med' saker liksom. Och då kanske man glömmer bort mer det melodiska.]

A similar view was expressed by Kristoffersson when working on our trio versions of Cortége and Ludus Tonalis:

JK: I imagine that it easily happens when you take a song where the melody is very much based on chord changes. And to then play solo on it, then you can become very locked in ‘the box’ somehow, when the theme is so connected to the harmonies. So, to open it up a bit in some places ... that was a good thing, I think. (Listening session with Kristoffersson & Olsson, 2023-11-04)
[JK: Jag kan tänka att det lätt blir så när man tar en låt där melodin bygger mycket på att det växlar ackord. Och att sen spela solo på det, då blir man ju väldigt låst i lådan på något vis, när temat är så ihopkopplat också med ackorden. Så att öppna upp det lite på sina ställen ... det var bra, tror jag.]

Choosing a more open interpretation of the harmony was something we had discussed when rehearsing with the trio, as a way of overcoming challenging chord progressions. The way we conceived this was to think more in terms of harmonic landmarks rather than individual chords, such as in this illustration of Cortège:

Figure 9.15. Chord progression for Cortège with harmonic landmarks.

Strategies of this type often turned out to be successful for achieving fluency in dense passages. Similarly, as I practiced the end of the solo section in Cortège, having a focus on the general direction and the intensity of the passage – rather than individual chords – made it easier to achieve a sense of momentum in the improvisation.

There were pieces, however, when this strategy didn’t prove to be as useful. This was the case, for instance, with Interludium: Romantic miniature from Ludus Tonalis, where the underlying tonalities were more fluctuating:

JK: I was probably quite restricted by the chords there, I felt. …  I thought it was more difficult on this song than on Cortège to sort of let go of each chord. …

PK: And the song is also in the original quite “twisted”, there are no really clear “home places”. Well, it’s in E in the beginning for a while, then we're in B, but in between it moves a lot. It feels almost random at times. (Listening session with Kristoffersson & Olsson, 2023-11-04) 

[JK: Jag var nog ganska styrd av ackorden där, kände jag. ... Jag tyckte det var svårare på den här låten än på Cortège att liksom släppa på varje ackord. …


PK: Och låten är ju också i original ganska så "skruvad", det är inga jättetydliga "hemmaplatser". Eller, det går i E i början ett tag, så är vi i B, men däremellan så rör det sig väldigt. Det känns nästan slumpmässigt ibland.]

Situations like these, then, seemed to call for other approaches.

Line-based approaches

Another way to handle intricate harmonies was to focus more on linear aspects and melodic continuity. This could involve finding lines over a particular chord progression, such as guide-tone lines that create linear connections between chords. I discussed this with Kristoffersson, in relation to improvising over D’un vieux jardin:

JK: I probably worked quite a lot on trying to find some kind of longer phrases. Maybe over 3, 4, or 5 bars, and be able to find a line ... 
And when the tempo is a little slower, it's also easier to be a bit more fluent in the phrases, in my experience. So, yes, I probably thought about longer, tried to make it work with longer phrases.

PK: But it's interesting, because I also worked quite a lot with lines, over 4 bars, for example, and "what kind of... is there an ascending line here, for example, that can link these strange chords together?". Or something to that effect.

JK: Yeah, exactly, and then it was also quite clear, as I heard now when we were listening, that.... if I ended up in the wrong place, that it was difficult to find my way back in, … it was more difficult to find those lines. (Listening session with Kristoffersson & Olsson, 2023-11-04)

[JK: Jag jobbade nog ganska mycket med att försöka hitta liksom längre fraser. Kanske över 3, eller 4, eller 5 takter och kunna hitta någon linje... 
Och när det är ett litet lugnare tempo är det ju också lättare att vara lite mer flytande i fraserna, upplever jag. Så att, ja, jag tänkte nog längre, försökte få till det med längre fraser.

PK: men det är intressant, för jag jobbade också ganska mycket med linjer, över 4 takter till exempel, och "vad finns det för någon sorts, finns det en uppåtgående linje till exempel här, som kan binda ihop de här konstiga ackorden?" Eller något sådant där.

JK: Ja men exakt, och då var det också ganska tydligt, hörde jag nu när vi lyssnade, att... om jag hamnade fel, att det var svårt att hitta in, … svårare att hitta de linjerna.]

The example below shows how I practiced a line-based approach to the chord progression of D’un vieux jardin, using a motif from the melody:

Figure 9.16. Example of line-based approach for D’un vieux jardin, mm. 7-11.

Extrapolating material from the melody

Using the melody was often useful in various ways when working with an improvised section, sometimes as a sort of “lifeline” when feeling momentarily lost, but also as a way of finding lines that could connect the different chords. At times, it was also helpful to work out elaborated versions of phrases from the melody to find new melodic possibilities for a given harmony progression. The illustration below shows a phrase from D’un vieux jardin, as an example of how I might use the melody phrase from measures 1-2 to create a sequence when practicing: 

Figure 9.17. Example of melody as basis for improvisation, based on D’un vieux jardin, mm. 1-4.

Internalizing sounds

Improvising over a particular progression typically became a question of getting it into one’s hearing, to internalize its sounds. As Evenstad pointed out, just having the information of a chord symbol was usually not enough:

TLE: Of course I look at the chords, and I might go through them, but it feels like it doesn't really sink in until you play together. Like, for instance, when I ask you; can you play this C#7 (#11), so I can play and take it in? Because then it's like it settles in a lot faster. Then it sinks in with me, when I get to do it that way. (Listening session with Evenstad, 2023-03-09)

[TLE: Det är klart att jag tittar på ackorden och kan gå igenom, men det känns som att det inte sätter sig förrän man spelar tillsammans. Som till exempel när jag frågar dig: kan du spela det här C#7 (#11), så att jag kan spela och känna in det? För då är det som att det sätter sig mycket snabbare. Då sätter det sig hos mig, när jag får göra det på det sättet.]

This was also the case when working with many of the other performers, that a repeated listening to the sounds of a particular progression was crucial for being able to improvise over it convincingly. As addressed under Audio-based approaches, listening and playing along with recordings of the music could be one way of achieving this.

Finding common notes

With harmonic progressions that contained a high degree of movement, was to search for common notes between different chords and scales. These connection points were valuable for my preparations. With some harmonic progressions, it also involved looking at connections between super-imposed triads, so-called “upper structures”.

Complex sonorities

At times, particular sonorities[1] seemed to pose certain challenges from an improvisational perspective, in particular structures that couldn’t be easily analyzed as familiar chords from a jazz improviser’s perspective.

Examples can be found in the following passage, from Dutilleux’s Sonatine for Flute and Piano, pt. II:

Figure 9.18. Sonorities from Sonatine for Flute and Piano, pt. II. Harmony and piano voicings based on the original piano part (Dutilleux, 1943).

As a general rule, I tried not to simplify these sonorities and, for instance, turn them into more common chords from a jazz performer’s perspective; e.g., turning a chord like C#m7/C into a symbol such as C7alt. The challenge, however, was to improvise over such passages without getting stuck in conceptual approaches, for instance, relying too much on the upper structures.

It was not uncommon that I returned to certain sonorities over time, to further examine what scale choices might be compatible with the sounds in question. One example of this was the sonority from measure 21 of Assez lent, a sound that initially made Söderqvist and me lose momentum when we first tried improvising over the piece:

Figure 9.19. F(#5)/E sonority from Valses Nobles et Sentimentales: II. Assez lent, mm. 19-24.

The musical events leading up to this shouldn’t be ignored; not only is this sound foreign to the perceived tonal center of the passage (Bb minor), it also stands out because of the fact that we’re coming from a passage with a faster harmonic rhythm. My initial strategy was to use the general tonal center of Bb minor, with the required adjustments, to find my way when improvising around the sonority:

One approach that I found … was to stay in the general key and have that as a basis, but using my ears to adapt to the notes that stand out, like an E in the bass, but still retaining the overall tonal center of that spot, which is Bb minor. (Audio journal, 2022-06-15)

However, as I returned to this passage later, I noticed that my ears gravitated more toward pitches that would continue the chord sequence established in the previous measures (mm. 17-22):

| Bb (mel. min) | Ab (mel. min) | Bb (mel. min) | Ab (mel. min) | Gb (mel. min) | –“– |

Somehow, the descending motion from the previous measures seemed to create a gravitational pull towards this sound, despite the clash that occurred between the pitches of this scale and the bass note (an E natural). Also, Gb melodic minor was a choice that worked well with the following chords – F(#5)/Eb and F7(#5) – which acts as dominant chords to the subsequent Bb minor.

I was also curious how other musicians would interpret this sound. When I discussed it with Eirik Hegdal, he mentioned how his ears were spontaneously drawn to a D minor melodic scale as I played the sonority (Conversation with Hegdal, 2024-05-09). This alternative might seem somewhat unexpected, as there is no D in the chord progression or in the bars leading up to it. But, as he explained it, there was something in the sound that seemed to evoke that association, probably based on a jazz improviser's usual way of hearing chords and associating them with certain pitches.

Thus, three different interpretations could provide departure points for the same sonority – Bb melodic minor, Gb melodic minor, and D melodic minor – which is not unusual with sounds that are based on symmetrical structures, such as augmented triads.

Rethinking traditional chord/scale relationships

There were also moments when working with the compositions, particularly with Takemitsu’s music, where I realized that I had to move away from traditional conceptions of chord functions and chord/scale relationships, simply because they often seem inadequate when approaching this type of musical language. There was often ­– from a jazz improviser’s ear – a sense that one or two pitches had been added to pitch collections that could be identified as a scale. In Burt’s (2001) explanation, not only did Takemitsu often work with building vertical structures using pitches derived from modal scales, he would also work with intensifying such pitch collections by adding ‘chromatic’ pitches external to the mode in question (p. 28). Examples of this can be found in this passage of arpeggios from All in Twilight I, where colors are emerging gradually over different chord arpeggios:

Figure 9.20. Arpeggios from All in Twilight I (Takemitsu, 1989), mm. 47-54, with added chord analysis.

In the second bar, there’s a combination of C minor major 7 chord with the notes of a F# minor triad appearing above:

Figure 9.21. Arpeggio from All in Twilight I, m. 48.

In the sixth bar of the same passage there is a descending arpeggio, where an E minor 9 chord is unfolded in a zig-zag pattern, but where there also is the note Bb – a flat fifth – one octave below the natural fifth, similar to a blue note.

Figure 9.22. Arpeggio from All in Twilight I, m. 52.

The approach for this passage was to find a combination of the underlying chords – as I perceived them – and the additional colors, mainly through listening.

Recording 9.19. Ending solo from All in Twilight I, Härenstam & Knudsen. From Kingside Studio, Gnesta, 2022-03-23.

A similar phenomenon happened with the music of Dutilleux, such as Sonatine for Flute and Piano pt. II, where I realized that it wouldn't work over the dominant chord to find one common scale to base it on, since the melodic phrase alternates between the natural 9 and the flat nine:

Figure 9.23. Sonatine for Flute and Piano, pt. II (based on Dutilleux, 1943)

This led me to open up for the possibility of alternating between various extensions over the same sonority, extensions that do not add up to what we would normally refer to as a scale. But it serves to illustrate that scale-based thinking could seemed limited in relation to some of the repertoire.

Another example of a sonority that seemed to demand “big ears” was this sonority used repeatedly in Ruth Crawford Seeger’s Leggiero:

Figure 9.24. Sonority from No. 8, Leggiero from Nine Preludes for piano (Crawford Seeger, 1928).

A possible chord analysis might look like: Db add 9 (no 3d, add #11, add #15), with the high #15 on top, adding a distinct sound to the sonority. It should be noted that #15 is not a common extension. Although the spelling b9 would yield the same result on a piano keyboard, this does not reflect how the extension is used in this context; Much like the piece in general, sonorities are built using combination of fifth intervals, making the #15 top seem relatively consonant thanks to the simultaneously sounding #11 a fifth below.

A similar sonority was found in Tailleferre’s Pastorale in D – in the form of an arpeggio, in the opening phrase:

Figure 9.25. Pastorale in D (Tailleferre, 1920), first measure.

In this context, the extension is part of a bitonal passage created from right-hand phrases based on G# minor pentatonic played over a left-hand accompaniment based on D major. The way Evenstad and I got into that particular sonority was simply by repeating the first phrase as a basis for improvisation – as we did in our added introduction – which gave us a chance to adapt to the sounds, as well as trying different ways of highlighting dissonance or consonance depending on the pitches we emphasized. Evenstad also pointed out that there is a certain mystical quality to such a chord which removes expectations, opening your ears to other possibilities.

Besides gradually getting accustomed to such sonorities through repetition, variation, and adaptation, I also found it valuable to use them in composition, e.g. writing passages that incorporated such sounds. This is further covered in Composing using musical language.

Several pieces in the repertoire had sonorities that could be analyzed as “polychords”– i.e. the combination of chords where one triad/tetrad is superimposed over another – such as seen in the example below:

Figure 9.26. Example of polychord, D major 7 over C major 7.

This chord is found in an arpeggiated form – in a descending motion – in Hindemith’s Interludium: Romantic miniature, where the arpeggio movement contributes to smoothen the dissonance between the upper and lower parts of the sonority. This particular sound was something I, somewhat spontaneously, used as a basis for an open introduction to the piece:

Recording 9.20. Improvised piano introduction to Interludium: Romantic miniature (released as Contemplation on the album Reimaginations). Recorded in Studio Epidemin, Gothenburg, 2023-10-20.