The Voice

The synthesised voice was as much a part of Stephen Hawking as his wheelchair: flat, robotic, unmistakably American-accented, and instantly recognisable around the world. The story of how it came to be, and why he never replaced it, is one of the more human chapters in his life.

Losing his speech

For years after his diagnosis, Hawking's natural speech grew steadily harder to understand, until only those closest to him could follow it. The decisive blow came in 1985, when he caught pneumonia on a trip to Geneva and underwent a tracheotomy to save his life. The operation ended his ability to speak entirely. For a time he could communicate only by having someone slowly point at letters on a card while he raised his eyebrows to select them, an agonisingly slow process for a mind that worked so fast.

The system

The solution came from assistive-technology pioneers in California. Hawking was given a computer program that let him pick words and phrases from a menu on a screen, which were then spoken aloud by a speech synthesiser. At first he made his selections with a handheld clicker. As his hand function declined, the input method changed: a small sensor mounted on his glasses detected tiny movements of a single cheek muscle, letting him choose letters and words one twitch at a time. Word-prediction software, refined over the years, helped him build sentences faster than the painstaking input alone would allow.

It was never quick. He composed at a rate of only a handful of words per minute, which meant every public sentence he delivered was the product of real effort and forethought. That slowness, it has been suggested, was part of what gave his pronouncements their characteristic economy and weight.

Why he never changed it

The voice itself came from early synthesiser hardware and had a distinctly American accent, an odd quality for a man from Oxford and Cambridge. Over the decades, far more natural and more British-sounding synthetic voices became available, and he was offered upgrades many times. He declined them all.

The reason was simple and rather touching: the voice had become his voice. The world knew him by it, and so did he. He felt it was now part of his identity, and he did not want to give it up, even though it had been built from technology that grew more antique with each passing year. Engineers maintained and even rebuilt the ageing system over time specifically to preserve that exact sound.

A sound that outlived him

By the end, the toolkit that ran his communication was modernised behind the scenes and even shared openly so that others with similar conditions could benefit. But the voice the public heard never changed. It featured in his lectures, his television appearances and his cameos, and at his memorial his synthesised words were broadcast towards a distant black hole. A voice assembled from 1980s circuitry had become one of the most famous sounds on Earth, and a symbol of the refusal to be silenced.

For the full engineering story, see the technology that let him work.