Teach Sum, Cheat Sum is a lighthearted exploration of mapping graphic shapes and cultural associations to musical gesture. Its two human performers play a game, drawing shapes of mustaches with musical sounds, working from a graphic score that includes images of famous mustaches. Meanwhile a 1980s-style voice synthesizer housed in a small box sits between them onstage, altering the course of their dialogue with its disembodied but vaguely human sounds. The synthesized voice is both foil and playmate to the violins, pulling their games in new directions.
This piece is dedicated to the memory of composer Pauline Oliveros, who left a deep impression upon me and many, many others who encountered her and her music.
This synthesizer sounds in this piece are executed live during the performance: Inside the synthesizer box, an Arduino microcontroller directs a Parallax Emic2 text-to-speech module. The Arduino generates its commands during performance, meaning that every performance of the work is unique. The structure of the piece is hard-coded into the Arduino software, so that the form of the piece is strictly followed; but at every point in the work the Arduino makes use of weighted probabilities to decide what commands to issue and when to send them, creating local "improvisation" by the synthesizer. The Emic2 accepts commands from the Arduino and then executes them as best it can, sending sound output to a speaker attached to the synthesizer unit.
The use of this Emic2 voice synthesizer module in this piece is purposely intended to problematize the relationship between humans and automatons. Today, it is not always possible to immediately identify whether one is communicating with a living human or a sophisticated piece of technology. Teach Sum, Cheat Sum brings this to the fore by hacking the voice synthesizer system to produce sounds that run counter to its commercial purpose of clearly speaking intelligible text. Most of Teach Sum, Cheat Sum asks the synthesizer to spout consonants unattached to vowels, which are rare in speech. In fact, it was difficult to produce the streams of unvoiced consonants and gibberish for the piece, since this is not an expected use of the unit. The commands to the Emic2 from the Arduino are often interpreted as errors by the Emic2, and thus part of the compositional process was a long trial-and-error process of learning how to send barrages of "wrong" instructions to the synthesizer without this resulting in a barrage of error messages, which the Emic2 speaks out loud as they happen. And in fact each performance of this piece is a kind of tight rope walk to see if, when, and how many error messages the Emic2 will speak before the piece ends.
This work was commissioned by the Japan-USA Musical Perspectives, Miyuki Ito, Director.
Premiere performance happened March 12, 2017 at the Tenri Cultural Institute, New York City, by the violin duo Miolina (Mioi Takeda and Lynn Bechtold, violins).