Phonetics and speaking machines: On the mechanical simulation of human speech in the 17th century

This paper shows that in the 17th century various attempts were made to build fully automatic speaking devices resembling those exhibited in the late 18th-century in France and Germany. Through the analysis of writings by well-known 17th-century scientists, and a document hitherto unknown in the history of phonetics and speech synthesis, an excerpt from La Science universelle (1667[1641]) of the French writer Charles Sorel (1599-1674), it is argued that engineers and scientists of the Baroque period have to be credited with the first model of multilingual text-to-speech synthesis engines using unlimited vocabulary.