Misunderstood: Duologue with a Broken Machine
“Misunderstood: Duologue with a Broken Machine” explores the relationship between two imperfect entities: an algorithmic translator and a human reader. A North American actor reads the text of a Polish citizenship form as translated by his smartphone, trying to make sense of what he is being told to understand. Human and machine are seen to struggle in consummating an informational transaction. Who is chiding whom as the translation exercise progresses? The embodied performance, eliding between echo, question, and challenge, evokes that of playing with a ouija board—easy answers are sought from thin air, but instead the user must seek meaning and structure through seemingly-arbitrary external cues. Despite the promise of instant information, the networked smartphone profoundly fails at its task, leaving its human interlocutor at the mercy of a frustratingly, beautifully inscrutable set of instructions that can ultimately change the shape of his identity.
“Misunderstood: Duologue with a Broken Machine” was performed with one camera for the benefit of another. The English-language text was iteratively generated using Google’s language-translation software, its output captured via screen-recording software operating on the same phone. The camera recording the performance was remote controlled. The audio track was recorded in a single take while watching a playback of the translation in real-time.
The performance consists of a single 24-minute shot, edited down to 21 minutes in order to eliminate the page turns and software freezes. The tools used to create the experience are only those found in the performer’s apartment, echoing how he learns languages on his own. The performer is seen to tire, due to the length of the single-take shot—the average modern cinema shot is shorter than three seconds in duration. Meanwhile, the algorithm also appears to run out of steam as the mobile device’s RAM fills up: words and phrases take longer to transform as the performance progresses.