False Self Reads Bangor Sitting Meditations

False Self Reads Bangor Sitting Meditations is a systems music composition that sets spoken word to a background of sine waves, drones, harmonies and arpeggios. The work explores the Buddhist notion of no fixed self, by systematically blending three guided meditations that were recorded by three esteemed meditation leaders into a new monologue, that passes through eight musical backdrops. The monologue is recited by my own voice and digital clone of my voice.

The composition is part of my False Self series. The series began in 2016. False Self compositions work with found materials, such as existing texts, video or music and subject this material to systematic processes to create something new. These composition explore themes of identity, authorship and language.

False Self Reads Bangor Sitting Meditations uses the therapeutic language of Buddhist mindfulness as material. Specifically the words spoken in the Mindfulness Audio Downloads by the meditation leaders at Bangor University's Centre For Mindfulness Research And Practice (CMRP). 

The CMRP Mindfulness Audio Downloads are a guided meditation streaming resource that UK based Clinical Psychologists and Counsellors frequently signpost their clients to. The CMRP divide their mindfulness audio resources into Body Scans, Sitting Meditations and Walking Meditations. All the words and phrases spoken throughout the project are systematically sourced from the CMRP Sitting Meditations.

The creation of False Self Reads Bangor Sitting Meditations proceeded through seven distinct stages:

1. Preparing the material

2. Table of future connections

3. Constrained writing

4. Composing the meditation backdrops

5. Vocal recording

6. Digitally cloning my voice

7. Arrangement, sequencing and mixing

Preparing the material

I began by downloading the six Sitting Meditation audio files that are hosted on the CMRP website. These meditations were devised and performed by the meditation leaders Sarah Silverton, Bethan Roberts, Eluned Gold, Jody Mardula, Cindy Cooper and Rebecca Crane. 

Riverside FM's transcription service was used to transcribe the audio files and conveniently group the text into minutes. I then sequentially numbered the minutes by hand (note the green loop on the left side of the page below).

This process was undertaken for all six of the CMRP Sitting Meditations. Each Sitting Meditation was then assigned a letter:

• A: Sarah Silverton

• B: Bethan Roberts

• C: Eluned Gold

• D: Jody Mardula

• E: Cindy Cooper

• F: Rebecca Crane

Three letters were randomly selected (D, F & C). These were the Sitting Meditations that were created and spoken by Jody Mardula, Rebecca Crane and Eluned Gold.

• Jody Mardula's Sitting Meditation has a duration of 16 minutes and comprises of 1150 words, that were divided by Riverside FM's transcription service into 76 groups of text / minutes (henceforth termed word-groups).

• Rebecca Crane's meditation has a duration of 41 minutes and comprises of 2870 words, that were divided into 122 word-groups.

• Eluned Gold's meditation has a duration of 38 minutes and comprises of 2120 words that were divided into 122 word-groups.

All of the selected meditations were divided into a minimum of 76 word-groups, therefore the number 76 was set as the limit for the system.

Table of future connections

The word-groups were then organised into a table. The table was populated by randomly selecting 76 word-groups from each of the three selected CMRP Sitting Meditations and then entered them into the table beneath the speakers name.

The table has 76 rows. Mardula's meditation had 76 word-groups, therefore all of her meditation was used as working material. Crane and Gold's meditations were both divided into 122 word-groups, therefore 46 of their word-groups were randomly discarded and not used going forwards.

Reading across row four of the table we can see that that working-set four features Mardula's word-group three, Crane's word-group 99, and Gold's word-group 91.

Constrained writing

The tabulated word-groups were then used as materials in a constrained writing task to compose the spoken word monologue that is the core of False Self Reads Bangor Sitting Meditations.

The five conditions for the constrained writing phase of the composition were:

1. Words and phrases can be selected from any of the three word-groups that form the working-set. For instance, row 4 of the table is working-set 4 and is made up of Mardula's word-group three, Crane's word-group 99, and Gold's word-group 91. These were: 

D,03 Mardula: Feeling the back of your neck lengthening slightly and your shoulders dropping, lowering gently.

F,99 Crane: The breath is always with you, always there to help settle and steady you. Reconnecting us again and again and again with the present moment.

C,91 Gold: So allowing yourself to observe thoughts rather than being caught up in thoughts.

2. All three word-groups must be used.

3. Words from a word-group must be uttered sequentially.

4. Words can be omitted.

5. Words from one word-group of the working-set can interrupt either of the other two word-groups at anytime; however words must be spoken in the sequential order of the parent word-group.

6. Punctuation can be used freely.

Completed examples

Three examples of completed constrained writings are shown below. Words highlighted in yellow were selected to be included in the line. The sentences highlighted in green are the completed constrained writings that feature as part of the monologue.

Composing the meditation backdrops

The spoken monologue is the principle element of the composition. The primary function of the eight meditation backdrops is to support the voice by creating a musical cushion for it to rest upon.

The following compositional restrictions were placed upon the backdrops:

1. The meditation backdrops were all composed to a uniform tempo of 56 beats per minute. This pacing was chosen because it has the sense of a resting heartbeat.

2. Each backdrop was allocated an underlying time signature. These included 4/4, 5/4, 9/8, 13/16 alternating bars of 7/8 & 8/8, alternating bars of 8/8 & 9/8, alternating bars of 9/8 & 11/8, and so on. The time signature lightly masks the common underlying tempo and impacts the structure of the compositions, however it is largely disguised from the listener.

3. Each of the eight backdrops was assigned a different drone pitch.

4. Each backdrop was allocated a different harmonic character.

5. Each backdrop had a selection of seven to nine of the 76 meditation randomly utterances allocated to it. The remaining utterances were discarded.

So for instance, Meditation 1, has a tempo of 56bpm, is built with alternating bars of 9/8 & 11/8, a drone pitch of 294Hz (D#4), a harmonic character of musical intervals of fourths and seconds; and has eight meditation utterances assigned to it.

The musical backdrops were composed with an E-MU E5000 Ultra sampler and a Roland D50 synthesiser. These were recorded and programmed using a Yamaha QY700 MIDI sequencer. The two sound sources were enhanced by Korg A1 and Yamaha SPX90ii effects processors. The musical results were recorded into a modern computer workstation.

Vocal recording

To prepare for the recording I spent approximately 20 hours listening to the CMRP meditation leaders speak, before beginning to rehearse and develop my own meditation voice.

I recorded my vocals in my Helix Branch studio. I set six sound baffles around myself to create a white cocoon. The simple recording chain was a CAD M179 microphone into a Cranborne Audio Camden EC1 preamp. 

Digitally cloning my voice

All of the vocal recordings were fed into ElevenLabs Voice Lab. ElevenLabs is a British-Polish software company that specialises in developing natural-sounding speech synthesis software. Their deep learning software was used to clone my voice.

I instructed their software to recite the entire monologue and recorded the results. I then randomly chose and removed one recording of my own voice from each of the eight phases of the meditation and replaced it with the model of my voice that was generated by the ElevenLabs machine learning algorithm.

Arrangement, sequencing and mixing

*Currently looking for a mix engineer to work with*