The Voice on the Paper

Someone sings, and a line trembles through the soot.

It is April 9, 1860, in Paris. The inventor Édouard-Léon Scott de Martinville has arranged a machine to catch the movements of sound. Air makes a membrane vibrate. A light stylus follows its movement, tracing a wavering path across paper blackened by smoke.

The song is Au clair de la lune.

As the paper moves, the voice acquires a shape. The line follows the movement of the vibrating mechanism. The performance takes seconds. Afterward there is a sheet of marked paper where there had been an event in the air.

Scott’s instrument cannot play it back.

He has made a device for studying sound by looking at it. The trace is an object for the eye. Listening to this particular sheet will require a future he has not built.

The singer finishes. The paper remains.

The sheet survives in an archive.

Elsewhere, sound recording becomes an industry. Voices are captured on cylinders and discs, on magnetic tape, in electronic memory. People grow accustomed to hearing someone who is somewhere else, or someone who is no longer alive. A familiar voice can come out of a box. Entire generations learn that this is ordinary.

Scott’s trace has to wait.

In 2008, a team working with the First Sounds project and Lawrence Berkeley National Laboratory turns high-resolution images of the old recordings into audible sound. The researchers follow the tracings, work on their distortions, and translate visible movement into a signal that modern equipment can play.

Through the noise comes singing.

Nearly a century and a half separates the two events: a voice moving the air in Paris, and the recovered recording moving the air around new listeners. The original vibrations have long since died away. These are fresh vibrations, produced by a different machine, in another world of rooms and ears.

The line has helped one moment reach another.

Almost nothing else about that April performance has come through with it.

The recording cannot restore the room. It cannot tell us everything the singer was thinking. It cannot bring a person back to answer a question. Even the voice requires careful interpretation, and the first reconstruction will turn out to contain a revealing mistake.

Yet a little of a vanished event is sufficient to cause something new. A listener hears a tune. A researcher revises an account of recording history. Someone hears a human voice from a year they had known only through silent things.

This book follows that long reach.

A footprint can outlast the foot. A cell can inherit a useful change from an ancestor it will never meet. An animal can avoid a danger because of yesterday. A written sentence can alter the life of a person born after its writer’s death.

Different mechanisms do this work. They will lead us through chemistry, memory, voices, books, cables, and machines that compose answers. What connects them is a strange opening in the present: something here can stand in a useful relationship to something that is absent.

The relationship can guide a life. It can also mislead one. Making a record survive and making it understood will prove to be different achievements.

For now, there is a blackened sheet, a line that somebody took care to preserve, and people who have learned how to ask it a new question.

A century and a half late, the song begins.