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

Human speech may have a universal transmission rate: 39 bits per second

sciencemag.org

Post reply on HN
Print view View original

231–240 of 263 posts

Re: Human speech may have a universal transmission rate: 39 bits per second

#231

Post by tralarpa »

vanderZwan wrote: post #228
ahicks wrote: post #157

Earlier quoted context omitted.

And as a followup to that--because really this is the weird part--some circles of blind people (including mine) talk faster between ourselves. That's not common, but it happens. I still sometimes have to remember that other people can't digest technical content at the rate I say it and remember to slow down. A good way to bring it out is to have me try to explain a technical concept that I understand really well. I h…

I kind of want to hear a "blind Danish" conversation now. For those unaware: Danish is phonetically one of the most difficult languages in the world, to the point where children who acquire Danish as their first language on average start speaking a few months later than children with any other first language. To clarify: speaking is not the same as understanding - toddlers are often capable of understanding language…

Actually, understanding seems to be affected, too! See:

https://www.cambridge.org/core/journals/journal-of-child-lan...

Re: Human speech may have a universal transmission rate: 39 bits per second

#232

Post by tomp »

ahicks wrote: post #154
ilamont wrote: post #15

But, he says, instead of being limited by how quickly we can process information by listening, we’re likely limited by how quickly we can gather our thoughts. That’s because, he says, the average person can listen to audio recordings sped up to about 120%—and still have no problems with comprehension. Some years ago I worked on an accessibility project for an app and website designed for people with disabilities. One…

I'm the blind dev who refactored a huge chunk of the Rust compiler [0]. I'm at roughly 800 words a minute with a synth, with the proven ability to top out at 1219. 800 or so is the norm among programmers. In order to get it we normally end up using older synths which sound way less natural because modern synthesis techniques can't go that fast. There's a trade-off between natural sounding and 500+ words a minute, and…

Has anyone tried overlapping words instead of speeding them up? Like so:

  How
   are
    you
     doing?
I often wondered if this, or at least sped up speech, should be the default robotic interface... it would make sense to optimize for efficiency/speed (while maintaining legibility) if we can do so.

Re: Human speech may have a universal transmission rate: 39 bits per second

#233

Post by kortex »

JohnBooty wrote: post #200
bradknowles wrote: post #96

Earlier quoted context omitted.

If you really want to understand it, I would encourage you to write down notes about it by hand. For some reason I do not yet understand, the motion of physically putting pen(cil) to paper helps ingrain that information into your brain, in a way that typing it into a computer does not.

IMHO it's primarily because you have to process and comprehend the information in order to make a decision about what's important enough to jot down. Although, anecdotally, I find that writing notes by hand is better for recall than typing. I'm not sure if this is because writing activates different parts of the brain, or simply because writing is slower and forces me to therefore think harder and comprehend better b…

I think this is likely it. In my opinion, it reminds me of autoencoders. Having to compress the information means understanding the content enough to discard unnecessary bits.

I can roughly keep up in real time with speaking on the keyboard. Even though speaking is a bit faster than my max rate around 90 wpm, by dropping filler words and gaps I can mostly avoid summarizing and quote verbatim. When transposing at full speed like that, I feel like a conduit at times, very little sticks deeply in memory.

Contrast with writing notes, where I'll only write down things I find particularly important. Most of the time I'm just trying to actively listen.

Purely anecdotally, I feel like when I am typing and summarizing to the same compression ratio as writing, my retention with typing is better, because I can do it without looking, it's faster, and I can tune back in to the speaker.

On the other other hand, with a laptop open, I'm much more likely to get distracted with emails/tasks.

Bottom line, I think we need to study more axes of this problem, if only because it gives neat insights into cognition.

Re: Human speech may have a universal transmission rate: 39 bits per second

#234

Post by tripzilch »

Claude Shannon (of the information theory fame) did a similar research with his 1951 paper "Prediction and Entropy of Printed English". He used a particularly clever idea, leaving out words or letters from English text, he then measured how accurately people could predict the missing text. And then used those numbers for statistical information-theoretical analysis to estimate the information density at about 9-11 bits (IIRC) per letter.

Re: Human speech may have a universal transmission rate: 39 bits per second

#235

Post by tripzilch »

ahicks wrote: post #154
ilamont wrote: post #15

But, he says, instead of being limited by how quickly we can process information by listening, we’re likely limited by how quickly we can gather our thoughts. That’s because, he says, the average person can listen to audio recordings sped up to about 120%—and still have no problems with comprehension. Some years ago I worked on an accessibility project for an app and website designed for people with disabilities. One…

I'm the blind dev who refactored a huge chunk of the Rust compiler [0]. I'm at roughly 800 words a minute with a synth, with the proven ability to top out at 1219. 800 or so is the norm among programmers. In order to get it we normally end up using older synths which sound way less natural because modern synthesis techniques can't go that fast. There's a trade-off between natural sounding and 500+ words a minute, and…

Are old speech synths not harsh on the ears to listen for longer periods? Or maybe I'm just familiar with the super robotic ones (I like them for music production).

If so, have you considered using an EQ plugin to maybe turn down the harsher high frequencies a few notches? Just a thought.

Re: Human speech may have a universal transmission rate: 39 bits per second

#236

Post by mizaru »

tripzilch wrote: post #234

Claude Shannon (of the information theory fame) did a similar research with his 1951 paper "Prediction and Entropy of Printed English". He used a particularly clever idea, leaving out words or letters from English text, he then measured how accurately people could predict the missing text. And then used those numbers for statistical information-theoretical analysis to estimate the information density at about 9-11 bi…

Looked it up, 11.82 bits per word, 2.62 bits per letter.

Re: Human speech may have a universal transmission rate: 39 bits per second

#238

Post by aepiepaey »

ahicks wrote: post #154
ilamont wrote: post #15

But, he says, instead of being limited by how quickly we can process information by listening, we’re likely limited by how quickly we can gather our thoughts. That’s because, he says, the average person can listen to audio recordings sped up to about 120%—and still have no problems with comprehension. Some years ago I worked on an accessibility project for an app and website designed for people with disabilities. One…

I'm the blind dev who refactored a huge chunk of the Rust compiler [0]. I'm at roughly 800 words a minute with a synth, with the proven ability to top out at 1219. 800 or so is the norm among programmers. In order to get it we normally end up using older synths which sound way less natural because modern synthesis techniques can't go that fast. There's a trade-off between natural sounding and 500+ words a minute, and…

> Whether and how much the skill transfers to normal human speech, or even between synths, is person-specific. I can't do Youtube at much beyond 2x. Others can. It's definitely a learned skill.

I find that the maximum understandable rate varies a lot between speakers. For some speakers 2.5x is possible, but just 1.5x for others.

One advantage synths has, is that they can more easily control the speed at which words are spoken, and the pauses between words independently. When watching/listening pre-recorded content I often find that I'd want to speed up the pauses more than the words (because speeding up everything until the pauses are sufficiently short make the words intelligible).

If someone knows of a program or algorithm that can play back audio/video using different rates for speech and silence, please share.

Re: Human speech may have a universal transmission rate: 39 bits per second

#239

Post by aepiepaey »

bradknowles wrote: post #96
dfhdshfff wrote: post #71

Earlier quoted context omitted.

I used to watch a lot of lectures at a high speed. But I've came to realize that the faster I watch something, the faster I forget it. It is like the infromation doesn't have the time to settle in my memory, despite me understanding it. It's maybe because when things are slow, I can use the dead time to think about the implication/corner cases of what's being said.

If you really want to understand it, I would encourage you to write down notes about it by hand. For some reason I do not yet understand, the motion of physically putting pen(cil) to paper helps ingrain that information into your brain, in a way that typing it into a computer does not.

Here's a paper on the topic which presents some ideas: https://journals.sagepub.com/doi/full/10.1177/09567976145245...

Re: Human speech may have a universal transmission rate: 39 bits per second

#240

Post by aepiepaey »

zamadatix wrote: post #39
gonehome wrote: post #30

Earlier quoted context omitted.

Is there a trick to get it above 2x (which is the cap in the UI). It'd be nice when watching videos of congress.

I do it via `document.getElementsByTagName("video")[0].playbackRate = 3.0` in the dev console. You can up arrow to it similar to a normal terminal on the next video. The nice thing about this is you can put any float in e.g. 3.333. There are also extensions of course.

You can also put this:

  javascript:(function(){document.querySelector('video').playbackRate=3.0;})();
as a bookmark on your bookmarks bar.
Post reply on HN