Factoring Speech into Linguistic Features

Spoken language technologies, such as automatic speech recognition and synthesis, typically treat speech as a string of "phones". In contrast, humans produce speech through a complex combination of se