Even though this machinery was limited in its abilities to process composite data in different layers due to its intricate engineering and design structure, it became a primitive intention to deep learning
Client | French Cultural Center Istanbul
In the year 2001, I was commissioned to create an art installation in memoriam of Marcel Proust for the French Institute Istanbul. The project was challenging for both art and design practices. But the scale and the potential were intriguing.
Marcel Proust had defined the sound of people's voice as a language in his book: In Search of Lost Time. “The sound of her voice, having never before noticed that it was so different from the voice of other people. I found them all unlike, molded in a language peculiar to each of the speakers, each playing upon a different instrument.” — Marcel Proust.
If the sound was peculiar to each speaker; as an idea, even generating sound arrays, and playing back them in an impulsive order would create a language. But how to obtain such spontaneity?
The sound was the tool for spontaneity. And it needed to be played by an instrument. The theremin was the best choice due to its similarity to the human voice and the ability to be played by motions. Once the tool embedded into the machinery where motions generated sounds, the sum of the experience would create a language for each user/audience.
The question was how to build the machinery? The first layer was installing a space: A cyclorama where there was no sense of space. The second layer was producing composite soundtracks with the applicator; a theremin. The third layer was connecting those tracks to motion sensors. So when an audience walked in the cyclorama, her motions activated sensors attached to the relevant tracks. And the fourth layer was, through headphones, playing back the tracks to the audience. In the end, this impulsive order that she navigated had created a sound array described as her language.
Even though this machinery was limited in its abilities to process composite data in different layers due to its intricate engineering and design structure, it became a primitive intention to deep learning.
