3 ms·

Sadly the last time I looked into this (though, I suppose I looked at roman pottery) I came to the conclusion that a) this is unlikely to have accidentally occu

Sadly the last time I looked into this (though, I suppose I looked at roman pottery) I came to the conclusion that a) this is unlikely to have accidentally occurred and b) their machining would have been unlikely to be up to reliably (re)producing instruments that were able to do this intentionally. Recording voice takes responsiveness to forces a tiny fraction of a percent of, say, a foot pedal controlling part of a lathe. I'm not saying it's impossible but such a feat intentionally constructed surely would itself be worth recording. EDIT: clarified tense

adastra22 2y ago
Maybe. Yet similar logic could be used to say that papyrus scrolls burned to a crisp would be unreadable. Yet here we are: https://www.neh.gov/news/students-decipher-2000-year-old-herculaneum-scrolls https://www.neh.gov/news/students-decipher-2000-year-old-her... I went with Egyptian rather than other ancient pottery cultures (there are many to pick from!) because of the mass production of identical figurines through repeated processes (allowing for differential comparison), and the use of reed tools. The foot-pedal signal would clearly dominate if you just played back a recovered signal from the reed on clay. As would also the heartbeat of the worker holding the reed. And of course all of the signal would be distorted by the drying process of the kiln. But from an information-theoretic standpoint, the information (might!) still be there, and decipherable if it can be scanned at high enough resolution and enough compute thrown at it, and these other sources of noise filtered / separated out. I hold out hope that this will one day be possible.
- darby_nine 2y ago
  > Yet similar logic could be used to say that papyrus scrolls burned to a crisp would be unreadable. We also know what papyrus scrolls look like, know what they were used for, and have a lot of examples of writing from that time period and place we can use to actively search for more, similar data. If you think about the mechanical process that must take place and the materials that they had at hand to record audio the chances of this happening accidentally in a way we could know how to read or look there without having witnessed the production of the audio to begin with are just staggeringly low. But, of course, I would love for this to be possible too, and I also imagine there's some really interesting stuff out there if one knows just where to look and how to read it. I had considered glass before as it's perturbed by the air as it's blown and holds its form well very quickly as it cools. But I wouldn't know the first thing of how to read that, and blown glass is much more delicate and rarely survives.
  - adastra22 2y ago
    > without having witnessed the production of the audio We know how they made pottery. We could recreate the process ourselves and look for that signal first (which we know is there because we put it there), before doing any invasive sampling of ancient artifacts.

adastra22 2y ago

Maybe. Yet similar logic could be used to say that papyrus scrolls burned to a crisp would be unreadable. Yet here we are: https://www.neh.gov/news/students-decipher-2000-year-old-herculaneum-scrolls https://www.neh.gov/news/students-decipher-2000-year-old-her... I went with Egyptian rather than other ancient pottery cultures (there are many to pick from!) because of the mass production of identical figurines through repeated processes (allowing for differential comparison), and the use of reed tools. The foot-pedal signal would clearly dominate if you just played back a recovered signal from the reed on clay. As would also the heartbeat of the worker holding the reed. And of course all of the signal would be distorted by the drying process of the kiln. But from an information-theoretic standpoint, the information (might!) still be there, and decipherable if it can be scanned at high enough resolution and enough compute thrown at it, and these other sources of noise filtered / separated out. I hold out hope that this will one day be possible.

darby_nine 2y ago
> Yet similar logic could be used to say that papyrus scrolls burned to a crisp would be unreadable. We also know what papyrus scrolls look like, know what they were used for, and have a lot of examples of writing from that time period and place we can use to actively search for more, similar data. If you think about the mechanical process that must take place and the materials that they had at hand to record audio the chances of this happening accidentally in a way we could know how to read or look there without having witnessed the production of the audio to begin with are just staggeringly low. But, of course, I would love for this to be possible too, and I also imagine there's some really interesting stuff out there if one knows just where to look and how to read it. I had considered glass before as it's perturbed by the air as it's blown and holds its form well very quickly as it cools. But I wouldn't know the first thing of how to read that, and blown glass is much more delicate and rarely survives.
- adastra22 2y ago
  > without having witnessed the production of the audio We know how they made pottery. We could recreate the process ourselves and look for that signal first (which we know is there because we put it there), before doing any invasive sampling of ancient artifacts.