The Diary

Sound and installation, dimension variable, 2017

180 diary entries from The Unabridged Journals of Sylvia Plath were recorded using Google text-to-speech technology(Google Translate). As text-to-speech service is more and more common in many fields, scientists dedicate to make its functioning better and closer to how human articulate language. What interests me particularly is the suprasegmental aspects of phonology— stress, intonation, pause, tone, which are closely related to complicated and subtle emotional expressions of human beings. How does artificial intelligence understand and pronounce words when they read? And how would they “interpret“ the texts?