Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Can text be made to sound more than just its words? (2022)

arxiv.org

Post reply on HN
Print view View original

21–30 of 32 posts

Re: Can text be made to sound more than just its words? (2022)

#21

Post by wincy »

andai wrote: post #17

Many moons ago I became quite obsessed with analyzing spectrograms on my computer. I would load up audio files in Audacity and look at them to see how the audio "looked", as a function of how intense each frequency is over time. You can even set a track to spectrogram while recording which allowed you to see the sound in real time. Music also tends to be very beautiful in the spectrogram! And birdsong also. Sometimes…

I knew a blind guy who did a trial where he could “see” using his tongue. Pretty neat!

https://news.wisc.edu/a-taste-of-vision-device-translates-fr...

Re: Can text be made to sound more than just its words? (2022)

#22

Post by eszed »

foofoo12 wrote: post #9

Very interesting idea. I remember reading that in visual spoken communications, only 20% is the actual words. The rest is tone of voice, body language, context, emphasis, expressions, ... all that stuff. I don't know if 20% is correct, but I feel it's very close to it. I also think a lot of internet arguments happen as a direct result of miscommunication. Emojis are great, but they get abused to the point that HN fil…

There's an acting exercise (it's from Joan Littlewood via Clive Barker) where one speaks "gibberish" - making language sounds, but not words - which, almost automatically, once they drop their terror of doing it, opens students up to all of those other avenues of communication. Later, you can switch students back and forth between the script and gibberish, and it becomes plain that if you can't play a scene as clearly (to those in it, not considering the audience) in gibberish as you can with words then you don't fully understand it.

Re: Can text be made to sound more than just its words? (2022)

#23

Post by pnut »

embedding-shape wrote: post #5
voxleone wrote: post #4

Emojis absolutely have their place here. They can add tone, nuance, and a bit of humanity where plain text can feel flat.

I feel like emojis is the lazy persons way of adding tone, nuance and humanity, when you don't know how to do so by only writing. Don't want to imply it's wrong, it's valid to be lazy, especially when it comes to improving communication, but I find myself thinking "How can I make sure this comes across as the joke it is?" and after one or two minute I just end up slapping a wink emoji at the end and don't rewrite the…

An idea compressed down into a single character is elegant and efficient.

Re: Can text be made to sound more than just its words? (2022)

#24

Post by kiicia »

andai wrote: post #17

Many moons ago I became quite obsessed with analyzing spectrograms on my computer. I would load up audio files in Audacity and look at them to see how the audio "looked", as a function of how intense each frequency is over time. You can even set a track to spectrogram while recording which allowed you to see the sound in real time. Music also tends to be very beautiful in the spectrogram! And birdsong also. Sometimes…

There was a guy who was able to recognize music just by looking at grooves of vinyl recording https://en.wikipedia.org/wiki/Arthur_Lintgen

Re: Can text be made to sound more than just its words? (2022)

#26

Post by ethmarks »

foofoo12 wrote: post #20
Isognoviastoma wrote: post #11

Earlier quoted context omitted.

Easy to check: try to speak with someone talking foreign language you don't know and estimate what percentage of what they said you understood from tone of voice etc. I would guess it's less than 80%.

That's very easy and very wrong. Let's say you have a 100 page book. Page 1 contains fundamental knowledge that allows you to understand the rest of it. If you skip page 1 then you won't understand the other 99. How much of the book will you understand if you only read page 1?

But tonal information can be parsed without lexical understanding and vice versa.

Somebody cursing in French can still be interpreted as anger even if you don't understand French, and written profanity can still be interpreted as anger even if you didn't hear it spoken.

Tone and language do complent each other, but neither is a prerequisite for the other like your book analogy would suggest.

Re: Can text be made to sound more than just its words? (2022)

#27

Post by foofoo12 »

ethmarks wrote: post #26
foofoo12 wrote: post #20

Earlier quoted context omitted.

That's very easy and very wrong. Let's say you have a 100 page book. Page 1 contains fundamental knowledge that allows you to understand the rest of it. If you skip page 1 then you won't understand the other 99. How much of the book will you understand if you only read page 1?

But tonal information can be parsed without lexical understanding and vice versa. Somebody cursing in French can still be interpreted as anger even if you don't understand French, and written profanity can still be interpreted as anger even if you didn't hear it spoken. Tone and language do complent each other, but neither is a prerequisite for the other like your book analogy would suggest.

> but tonal information can be parsed without lexical understanding

Parsed perhaps, but it's so context sensitive that it's not useful, save for extremities. The same tone of voice can have so many meanings based on what's actually being said and yet another if you add context.

Re: Can text be made to sound more than just its words? (2022)

#28

Post by kalavan »

foofoo12 wrote: post #20
Isognoviastoma wrote: post #11

Earlier quoted context omitted.

Easy to check: try to speak with someone talking foreign language you don't know and estimate what percentage of what they said you understood from tone of voice etc. I would guess it's less than 80%.

That's very easy and very wrong. Let's say you have a 100 page book. Page 1 contains fundamental knowledge that allows you to understand the rest of it. If you skip page 1 then you won't understand the other 99. How much of the book will you understand if you only read page 1?

That then raises the question: what is a unit of communication?

If communication is 20% verbal and 80% nonverbal, and if communication is very nonlinear in understanding (as with your book example), how do we know what 1% of communication is? What does it mean, and how can we tell that the figure is correct, when our main or only way of detecting whether communication succeeded is through understanding or lack thereof?

Re: Can text be made to sound more than just its words? (2022)

#29

Post by foofoo12 »

kalavan wrote: post #28
foofoo12 wrote: post #20

Earlier quoted context omitted.

That's very easy and very wrong. Let's say you have a 100 page book. Page 1 contains fundamental knowledge that allows you to understand the rest of it. If you skip page 1 then you won't understand the other 99. How much of the book will you understand if you only read page 1?

That then raises the question: what is a unit of communication? If communication is 20% verbal and 80% nonverbal, and if communication is very nonlinear in understanding (as with your book example), how do we know what 1% of communication is? What does it mean, and how can we tell that the figure is correct, when our main or only way of detecting whether communication succeeded is through understanding or lack thereo…

> when our main or only way of detecting whether communication succeeded is through understanding or lack thereof

That's not even a good test, due to miscommunication. Both parties might think it succeeded, but then much later on you find out the truth (maybe).

Re: Can text be made to sound more than just its words? (2022)

#30

Post by m463 »

kiicia wrote: post #24
andai wrote: post #17

Many moons ago I became quite obsessed with analyzing spectrograms on my computer. I would load up audio files in Audacity and look at them to see how the audio "looked", as a function of how intense each frequency is over time. You can even set a track to spectrogram while recording which allowed you to see the sound in real time. Music also tends to be very beautiful in the spectrogram! And birdsong also. Sometimes…

There was a guy who was able to recognize music just by looking at grooves of vinyl recording https://en.wikipedia.org/wiki/Arthur_Lintgen

I remember being able to recognize one song on vinyl.

It was a (telarc I think?) recording of the 1812 overture.

The grooves were wide where the canons went off, so that the needle could deflect enough to capture the dynamic range. You could see the waveform.

I think of "Surely You're Joking Mr. Feynman" where people could sniff like a bloodhound. Feynman would have people handle books, and he could tell which ones had been handled.

I think there are things that just trying would be successful more than you think.

Post reply on HN