A brief history of synthetic speech

The use of speech generating systems such as that which Stephen Hawking used have a history that stretches back further than most expect, over 200 years in fact. Attempts to produce synthetic speech was pushed initially by a desire to study how the voice functioned. It was only later that scientists looked to it as a means to aid those who couldn’t speak.

The first attempt to create a speaking machine was in the late 1770s. Russian professor Christian Kratzenstein succeeded in building a machine that produced vowel sounds using resonance tubes that were connected to organ pipes. It was only a few years later that Wolfgang von Kempelen, a scientist in Vienna, built his Acoustic Mechanical speech Machine, which was able to produce not only singe sounds but combination sounds and even full words. The machine consisted of a pressure chamber for the lungs and a leather tube for the vocal tract. Vowels were created by manipulating the leather tube and consonants by the use of four separate constricted passages. For plosives he created a movable hinged tongue and pair of movable lips.

The 1930s gave way to the first electronic speech machine, created by acoustic engineer Homer Dudley. The machine was an advancement on a machine used during World War I as a way of sending encrypted messages, encoding the speech input at one end and recreating it at the other. The Voder, first shown at the 1939 New York World Fair, was capable of creating speech without the initial human input. Instead speech was created by use of 10 keys, a wrist panel and a pedal. These would create electrical noises that were then manipulated to create words.

However, this was again proof of concept rather than something that was meant to be used by the public at large. It wasn’t until the 1960s that machines for public use started to appear. These early devices were mostly modified typewriters such as a sip-and-puff typewriter. The technological innovation during the 70s and 80s is what started to create products that are more familiar to us today. These were devices that produced a synthetic voice and controlled by systems such as Target Scanning, such as eye movement that is common today.

The 90s saw this technology develop and allowed the computer-based production of communication boards and devices begun to grow smaller and more portable, meaning the user was able to communicate wherever they were.