The always-interesting Laputan Logic site has recently posted an illustrated account of the Amazing Talking Head that Joseph Faber exhibited in Philadelphia in 1845. If this sort of thing interests you (as it should!), you can find on the Haskins Labs web site a series of pages on the early history of talking machines, which shows diagrams and/or descriptions of Kratzenstein's vowel resonators (produced around 1770 at the Imperial Academy in St. Petersburg), Wheatstone's version of von Kempelen's 1791 talking machine, and Alexander and Melville Bell's human vocal tract model, inspired by Wheatstone's machine and built during their boyhood in early 19th-century Edinburgh.
This is part of a larger Haskins exhibit entitled Talking Heads, authored by Phil Rubin and Eric Vatikiotis-Bateson, which also covers the history of electronic speech synthesis, virtual vocal tracts, models of speech production, the McGurk effect (we linked to that one earlier), facial animation, avatars, and more! There is an enormous depth of information on this marvelous site.
Another excellent site about speech synthesis from 1770 to 1970 can be found here, created by Hartmut Traunmüller at Stockholm University. Hartmut's page includes photographs of von Kempelen's original speaking machine, now in the Deutsches Museum in Munich.
Wolfgang von Kempelen was also the creator of the Turk, a chess-playing automaton that prefigured the recent Chatnannies algorithm developed by Jim Wightman. The Turk was actually operated by a smallish human concealed inside its cabinet -- some 15 chess experts and masters apparently played this role over the course of some 70 years -- but things moved more slowly in those 18th-century pre-internet days, and so the Turk played exhibitions all across Europe for several years before first being exposed. 50 years later, it was still successfully touring the U.S., where its popularity is said to have inspired the formation of the first American chess club in Philadelphia in 1826.
Unlike the Turk, von Kempelen's speech synthesis machine was a work of engineering with no hidden tricks, but (perhaps for that reason) it did not earn its inventor or its subsequent owners as much fame or money. However, one can argue that it played a part in the invention of the telephone a century later.
Posted by Mark Liberman at April 19, 2004 08:42 PM