My first poem uses “Green Light” by Lorde and Word2Vec to create dissonance and replace words. I decided to use a Lorde song because she is one of my favorite songwriters, and I was curious to see how Word2Vec would alter her lyrics (spoiler: it made them nonsensical!). I played with multiple bowed instruments to generate chords and extra random notes whenever I was replacing a lyric. As an added element of fun, pressing the right shift key plays a kick drum while the left shift key plays a hi-hat for wannabe drummers.
Code: https://drive.google.com/file/d/1qveXXS-3eYnIANiKOegw5zLemmXzwLKP/view?usp=sharing
My second poem takes user input to create a love story between two celebrities. I wanted to play around with sporking to “blend” asynchronous notes together to create chords that transitioned from a jovial mood to a somber one, with some added dissonance. As opposed to the first poem which has a mostly pre-defined melody, this poem randomly plays notes from chords (to which I added some extra minor 2nd intervals). Pressing the space bar replaces words using Word2Vec, again allowing an element of randomness.
Overall, I thoroughly enjoyed this assignment as a way of learning how to use ChucK. Having grown up learning music theory, it was hard to surrender to using random notes and words, but the time I put into listening and re-listening to my poems almost made the random notes sound normal (aka it’s opened my eyes to accept a different kind of “generative” music).
Code: https://drive.google.com/file/d/1WOL2x8jETnz_Ic3NWYFgu7305pFZ_w4h/view?usp=sharing
Acknowledgements: MuseScore (https://musescore.com/user/10579646/scores/3747176), STK examples, ChatGPT for poem generation, Ge and Andrew in Discord
Reflection
Does music need to be deterministic to be good? Does randomness result in chaos or constant innovation?
Given my love for listening to songs as background music throughout my day (as Spotify claims — the “soundtrack to [my] life”), I used ChucK to emulate chords and melodies. When listening to the poems, I often found myself excited to see how words and melodies would be altered based on whatever word Word2Vec returned and consequent dissonance was added. This caused me to reflect — would I ever be able to passively listen to music if it were generative in nature? What if artists released songs that were somehow different with every listen? Could these become commonplace in the music industry? Personally, I’m rarely ever fully concentrating on a song I’m listening to. Something about knowing exactly how something is going to pan out is comforting; it’s why people have comfort movies or TV shows.
That being said, I would still call the two programs “art”. The poems express feelings and tell stories, qualities I usually think about when interacting with art. I believe the poems making “sense” is secondary to the overall “feeling” they create. What concerns me is how little creativity I had to use to generate them. I’m hesitant to consider every new iteration of the poems “innovation”. Most of my thought process involved determining where adding randomness would be “interesting” rather than “good” (if “good” can even be considered measurable). There is beauty in composing melodies, chords, and lyrics, and I’m not sure whether randomly generating aspects of songs takes away from this.
