Working with storytellling in Berio's Sequenza

The excerpt from the video documentation, shows me working on a falling gesture from the first page of Berio's Sequenza. I try to make use of the Wilhelm Larsson intonation contour pattern for phrase endings (marked with a yellow No. 3 in the score). 

In the video, I am also discussing with myself (in Swedish) that it makes sense to make it sound like an ending, and it is actually beautiful in that way. 

At this stage, I presented the process on understanding storytelling as musical qualities, on an online seminar at Karlstad University. 

(Presentation in Swedish)

A first analysis: Listening thoroughly

By analysing short parts of the voice recording containing Wilhelm Larsson’s speech, I set out to look for overarching musical qualities in his rhetorical means and prosodic patterns. Could favourite intervals, recurring rhythmic patterns, or perhaps even tonal structures that were maintained throughout a story be identified? 

Speech in music notation

This analysis was made by ear and translated into standard musical notation. Hereby, I had to limit the material to characteristic shorter selections of the recording, since this analytical method was very time-consuming. Already at the outset, I was well aware that a machine analysis of the data would be more efficient, and allow for the use of bigger data. However, I chose to work from my own listening, since I wanted to use the analysis as a way of developing my individual performer's voice. This also allowed me to select material that I found musically interesting. My analysis was based on repeated listening, until I found a certain phrase or part of a sentence that caught my ear, and stopped there. By looping that part, the melodic content would eventually stand out, and was then transcribed into music notation.

When a few of these transcriptions were created, I could identify that Wilhelm Larsson's storytelling had some recurring patterns, related to rhythm, melody and timing.

As examples of the musical structures identified, Figure 2 and 3 shows how sentences would sometimes be structured to remain in one key, and audio files 2 and 3 plays Wilhelm's voice:

Applying storytelling on two flute works

At this stage in the process, the flute playing experiments started. Now I tried to fully engage in an openminded explorative research for how to annotate my scores of Berio's and Bozzas solo flute pieces, using Wilhelm Larsson's storytelling tools. In these experiments, I used both a reflective journal and video documentation to catch the process.

Following up on Delattre [tbc]

Having studied the intonation contours by Delattre, I decided to make my own model of Wilhelm Larsson's intonation contours, based on his storytelling tools found in the first analysis. 

- Making use of the music notation system

- Showing pitch, time, phrases, breathing

- Instead of Delattre's numbering, I decided that a combination of colours and numbers would be easier for me to code when performing an score later on in the process. 

- A recurring "unwanted element" from listening to the voice recording was the noises from Wilhelm  Larsson clearing his throat. I decided to not exclude that from the analysis, but give these sounds their own symbol in the analysis (the turquoise colored box). 

All marked in yellow - has to do with pitch and melody

All marked in green - also pitchrelated, but more long notes

All marked in red - is about how Wilhelm gives stress to a sentence

Extras that I added - the blue box for Wilhelm changing his dialect, the breath sounds, and an arrow for enjambements (going straight into the next phrase)

The next step was to annotate all the voice recording in the same manner, and  adding my version of Delattre's intonation contour to the analysis. 

This could look like in the pdf to the right.

Delattre's intonation contours [tbc]

To conclude and collect my findings of the musical rhetorics in Wilhelm Larsson's story-telling, I decided to draw from the linguist Pierre Delattre, who already in 1966...

• Linguist Pierre Delattre ....French intonation contours in 1966 
• Clearly takes off from the idea of musical notation
• Although Delattre's analysis is made upon French language, I decided to use the idea to build my own model of "Wilhelm Larsson's intonation contours". 

Musical rhetorics

Moreover, the thorough listening revealed other rhetoric means that Wilhelm Larsson uses, and are closely related to musical gestures:

Wilhelm uses short pauses just before the most important part of a sentence (Audio file 4)

He uses different pitches to sonically imitate or represent other characters in the story (Audio file 5)

When telling stories about something humorous, Wilhelm uses larger intervals within the sentence (Audio file 6)

When pausing to think he often rests on a certain pitch (Audio file 7)

Annotating and analysing Wilhelm Larsson's speech

Earlier on in the process, I had tried machine analysis to convert Wilhelm Larsson's speech into musical notation. 

In my case, my attempts of using computer analysis was never fruitful, probably due to my own technical shortcomings. By first analyzing the sound file in Logic Pro, I could make an XML file that Sibelius could turn into a score. Nevertheless, I found it hard to use that score. Some of the background noise always ended up as notes in the score, which was not my intention. Furthermore, the Sibelius score also brought a tedious process of adding and moving barlines. 

Instead, I decided to stick with the more graphic representation of speech I had used also in my own attempt on Delattre's intonation contours. By doing so, I found Praat (+ ref. ) the most helpful way of analysing speech, and including the text. In Praat, I had to make several experiments on the pitch analysis settings, before finding the best way to graphically show Wilhelm's voice.

Content

• A first analysis: Listening thoroughly
• Speech in music notation
• Musical rhetorics
• Delattre's intonation contours
• Following up on Delattre
• Annotating and analysing Wilhelm Larsson's storytelling
• Applying storytelling on two flute works
• Recording the finalized versions

Analysis:

some annotations are more common then others. During this process I realized I could have made better use of the intonation contours of Delattre, like the "question" phrase ending with a big upward interval (x.x in figure x).

Perhaps this was due to the situation of Wilhelm Larsson's recording, where it is actually not a dialogue, but a monologue, and whenever he raises a question it is more of a rhetoric question. Thus, my analysis shows that I have marked such passages with the "showing empathy" symbol, or on other occasions, the "citing other people" symbol. 

Annotating Berio's Sequenza

Semi-graphic score

Short phrases - almost speech-like. Lends itself well for resemblances with story-telling. Also difficult for most flute players to "find meaning" in the interpretation, since many of us are so much more used to play more classic melodies. 

Builds on a lot of small pauses - the score has got a lot of red "pause-markings"

Yellow marks for phrase-endings also frequent

Annotating in this way helped me understanding the score. 

After a while I had to take away some of all the red markings, I could not keep everything in mind while playing. 

Annotating Bozza's Image

Working with storytelling in Bozza's Image

Here, from the annotation process of the Bozza score. I am working on the enjambement, and reasoning on how I can make that more as speaking. 

Many big interval jumps - where I have two codes to choose between, either humorous or empathetic. Have to reason with myself on which one is relevant for each phrase in the score.