Sound Archaeology

The Dawn of Recorded Sound in America

JAN 22 2012, 9:27 AM ET

We can now hear, across a gulf of 140 years, some silly noises and a count to six, one of the earliest audio recordings.


Until recently, the oldest recorded sounds of known date which anyone could hear had been captured in 1888 on the “perfected” phonograph introduced that year by Thomas Edison. But Edison had invented his original phonograph eleven years before that, in 1877–and recorded sound itself is even older: In the 1850s, Édouard-Léon Scott de Martinville of Paris created the phonautograph, an instrument which scratched records of aerial sound waves on soot-blackened paper, not for playback, but for visual study. This means there is a big disparity between when sound was first recorded (around 1857) and the earliest recorded sounds we could actually listen to (1888).

That changed in 2008 when released a sound file created from a phonautogram of “Au Clair de la Lune” as sung on April 9, 1860. Suddenly we could hear more distantly into the past than ever before.

Even so, the intervening history of recorded sound — including the transformation by American inventors of the phonautograph into a “talking machine” — has remained frustratingly silent. The indented tinfoil sheets produced by Edison’s exhibition phonographs of 1877-78 weren’t regarded as permanently playable recordings, and little care was taken to preserve them in a playable state. No intelligible sound recovered from a historical tinfoil recording has ever been published.

So what else exists from before 1888? If we exclude recordings that weren’t intended for playback or to be permanently playable, then the oldest sound recordings preserved today are found at the National Museum of American History — experimental phonograms made starting in 1881 by the Volta Laboratory Association, which consisted of telephone pioneer Alexander Graham Bell, scientific instrument maker Charles Sumner Tainter, and chemist Chichester A. Bell.

With the support of a Lemelson Center Fellowship and the help of curator Carlene Stephens, I carried out a study of early sound recordings at the Museum, including the Volta materials, between October and December 2011. By comparing artifacts from the Volta collection with experiments described in notebooks at the Museum and the Library of Congress, I was able to identify a number of unlabeled items. One of these — a small copper disc with a laterally modulated or “zig-zag” sound recording — turned out to have been prepared shortly before October 20, 1881, to test whether electrically depositing a layer of metal on a recorded wax disc, and then using the metal negative to stamp out copies, might work as a basis for duplication in a future recording industry. “In this way a piece of music, for instance, can be recorded once,” Tainter had speculated, “and any number of copies made from this original record, and the music reproduced from each of the copies.”


The October 1881 date makes this one of the oldest known American sound recordings in existence, so a question naturally arises: What’s on it?  The written documentation I could find identifies the disc’s content only vaguely as “words and sounds … shouted into the mouth-piece,” but the Volta group’s notebooks reveal the general kind of test recitation they were then using, as for example:

July 4, 1881:  “Several trilled R’s–then–‘Mary had a little lamb, whose fleece was white as snow, and every where that Mary went the lamb was sure to go.’–Several trilled R’s–then–‘How is that for high’–trilled R’s–and–One–two–three–four–five–six–seven–eight–nine–”

July 9, 1881:  “There was a girl named O’Brian / Whose feet were like those of Orion, / To the circus she would go, / To see the great show, / And scratch the left ear of the lion. Trilled R’s. – ‘How is that for high’ more trilled R’s.”

Apart from the recurring expression “How is that for high” — roughly equivalent in 1881 to “How do you like them apples” — the most striking common denominator here is the “trilled R’s.” From laboratory notes, I could tell that this sound had recorded unusually well, and that the Volta group had often inserted it at beginnings, ends, and section breaks. But I didn’t know quite what it had sounded like. After all, nobody alive today had ever heard any of these experiments.


In December 2011, Dr. Carl Haber unveiled the first sounds extracted from Volta recordings using an optical scanning technology developed at Lawrence Berkeley National Laboratory in collaboration with the Library of Congress. One of the items chosen for this pilot project was the copper disc from October 1881. So now, at last, we can hear, across a gulf of 130 years: Trilled R’s–one, two, three, four, five, six–trilled R’s.

This post also appears on the National Museum of American History’s O Say Can You See? blog, anAtlantic partner site.