Clair de Lune

The most interesting article in the Times yesterday was surely the announcement of the discovery, in France, of a version of a sound recording made in 1860, some two decades before the first Edison phonograph. The recording was made on a phonoautograph, consisting of a recording horn and a stylus, which made an impression on a soft surface, much like an early wax cylinder. Although the recordings made on a phonoautograph were only intended to be looked at and studied as the visual representation of sound, using modern recording technology, the groves and indentations can be played back, and a somewhat ghostly voice can be heard singing the French folk song “Clair de Lune.”

I have long been fascinated by the oldest generation of recordings; voice recordings made by Sir Arthur Sullivan and William Gladstone in the 1880s, a horrible sounding piano roll, with a few barely audible notes containing all we have of Brahms at the piano; the recordings of Adelina Patti, the most famous soprano in the second half of the 19th century, with barely any voice left, but enough to get some sense of the former splendor of the voice, the earliest brass band recordings from the early 1890s, and so on, all with the crackle and distortion that, for instance, gives daguerreotypes their charm and fascination. In listening to old recordings on hears not only a particular recording, but the birth of an entirely new way of listening to the world, emerging in its dim, noisy, static-filled birth cry.

The discovery of this recording just underlines for me one of the great frustrations in the history of technology. Why oh why wasn’t the phonograph invented and developed earlier? As I understand the technological issues, there was absolutely no reason by the 1840s, by the time of the invention of the telegraph, that the best tinkerers in Europe or America couldn’t have invented the prototype of a recording device. Give it a decade or so to get developed, and by the 1860s we could have recordings conducted by Verdi or Wagner, Liszt or Clara Schumann at the piano; recordings of slave spirituals made during or immediately after the Civil War, and then somewhat later, blues recordings from the 1890s, or Buddy Bolden on the cornet. And old recordings take one back further than recording date—recordings made around 1900, like those of Patti, give evidence of what voices sounded like a generation earlier, so in listening to Patti, one hears operatic technique c. 1860. And if sound recording was truly possible in 1860, one could hear the voices of people born around 1800, and who played with Beethoven, or a few arias by Jenny Lind, or a few songs by Dan Emmett and other pioneers of the minstrel stage, perhaps accompanied by Stephen Foster on the piano.

There was an a fifty or sixty year period, from about 1840 to 1890, when it was possible to record images but not sound, and I have long wondered what it was like during this period of recorded asymmetry, when sounds were still assumed to be completely ephemeral and transient, while images were increasingly seen as something that could be locked and fixed. But somehow the transition from painting to photography seems, as epochal as it was, seems less astounding from what happened in 1860, when sound went from being forever ephemeral and transient, fading and lost forever almost immediately, to something that could be captured, controlled, and, eventually, reheard.