In my post “No, Newspeak Won’t Make You Stupid”, I explored the thesis that ‘cadence of information is constant’, that even if someone uses words which communicate more information, they will have to slow down their speech to compensate, thereby preventing them from communicating a larger amount of information using a rich vocabulary. I then present an alternative hypothesis for why we use rich vocabularies anyways.
One important crux of the thesis, is that the human mind is only able to encode and decode a certain amount of information per unit time, and that this limit is close to the speed we already naturally achieve. But on first reflection, this seems clearly incorrect—in the post, I calculate the cadence of information of speech to be a little more than 60 bits per second, which while a good amount, is a tiny fraction of the amount of processing the human brain actually does every second [citation needed] - surely we could speed up the rate we speak / listen, and still be able to keep track of what’s going on, right?
Well, in some sense, we probably could. But there are two reasons why I think this might not be perfectly right—for one, the parts of our brain that handle speech are only one part of many parts of our brain, with most of our information processing capacity not being designed to process speech, and there’s also the matter of what Simler and Hanson call “The Elephant in our Brain”, the ‘press agent’ part of our brain that cleans up our output to make us look like much better people than we actually are.
It’s not clear to me how much the fact that only a fraction of our information capacity is suited for handling speech affects this, but I do suspect that the “Elephant in the Brain” does have a substantial effect slowing down how fast we speak, making us talk slower so as to avoid revealing thoughts that may reveal intentions we don’t want revealed.
For this reason, both making people speak faster than they are comfortable with, and getting them drunk (effectively dilating time for them), are good ways to sus out whether someone’s intentions are pure—the second of these is captured by the phrase “In vino veritas”.
I regularly listen to lectures on YouTube at 2x speed and my comprehension is fine. I can even go a bit faster with a browser plugin. This took practice. I gradually increased the speed over time. I think I can probably read text even faster than that.
There are limitations. If the speaker has a thick accent, I can’t comprehend it as fast. If the concepts are difficult, then even though I understand the words, I often still have to pause the video while I think it through.
I have heard of blind people who use a narration interface on their smartphones and work up to 7x speed narration. If there are no surprises, humans can process language much faster than they can speak it.
Yeah, it definitely does seem to be possible to listen faster than we usually speak; at the same time, in public speaking classes, one is encouraged to speak slowly, so as to maximize the understanding of the audience. As you mention, the difficulty of concepts can require slowing down. While you can easily pause or rewind a video when you hit a part that you find tricky, you can’t do that in a presentation. Furthermore, what one person finds easy could be hard for someone else, but then the two people could be in the opposite positions a few sentences later.
Perhaps most ideas have some fraction of the people that need to process the idea, and the ideas that are hard vary from person to person, so in order to allow everybody to digest the parts they find tricky, a public speaker has to speak much slower than the people can actually understand, so no-one gets left too far behind while they’re deep in thought.
In my post “No, Newspeak Won’t Make You Stupid”, I explored the thesis that ‘cadence of information is constant’, that even if someone uses words which communicate more information, they will have to slow down their speech to compensate, thereby preventing them from communicating a larger amount of information using a rich vocabulary. I then present an alternative hypothesis for why we use rich vocabularies anyways.
One important crux of the thesis, is that the human mind is only able to encode and decode a certain amount of information per unit time, and that this limit is close to the speed we already naturally achieve. But on first reflection, this seems clearly incorrect—in the post, I calculate the cadence of information of speech to be a little more than 60 bits per second, which while a good amount, is a tiny fraction of the amount of processing the human brain actually does every second [citation needed] - surely we could speed up the rate we speak / listen, and still be able to keep track of what’s going on, right?
Well, in some sense, we probably could. But there are two reasons why I think this might not be perfectly right—for one, the parts of our brain that handle speech are only one part of many parts of our brain, with most of our information processing capacity not being designed to process speech, and there’s also the matter of what Simler and Hanson call “The Elephant in our Brain”, the ‘press agent’ part of our brain that cleans up our output to make us look like much better people than we actually are.
It’s not clear to me how much the fact that only a fraction of our information capacity is suited for handling speech affects this, but I do suspect that the “Elephant in the Brain” does have a substantial effect slowing down how fast we speak, making us talk slower so as to avoid revealing thoughts that may reveal intentions we don’t want revealed.
For this reason, both making people speak faster than they are comfortable with, and getting them drunk (effectively dilating time for them), are good ways to sus out whether someone’s intentions are pure—the second of these is captured by the phrase “In vino veritas”.
I regularly listen to lectures on YouTube at 2x speed and my comprehension is fine. I can even go a bit faster with a browser plugin. This took practice. I gradually increased the speed over time. I think I can probably read text even faster than that.
There are limitations. If the speaker has a thick accent, I can’t comprehend it as fast. If the concepts are difficult, then even though I understand the words, I often still have to pause the video while I think it through.
I have heard of blind people who use a narration interface on their smartphones and work up to 7x speed narration. If there are no surprises, humans can process language much faster than they can speak it.
Yeah, it definitely does seem to be possible to listen faster than we usually speak; at the same time, in public speaking classes, one is encouraged to speak slowly, so as to maximize the understanding of the audience. As you mention, the difficulty of concepts can require slowing down. While you can easily pause or rewind a video when you hit a part that you find tricky, you can’t do that in a presentation. Furthermore, what one person finds easy could be hard for someone else, but then the two people could be in the opposite positions a few sentences later.
Perhaps most ideas have some fraction of the people that need to process the idea, and the ideas that are hard vary from person to person, so in order to allow everybody to digest the parts they find tricky, a public speaker has to speak much slower than the people can actually understand, so no-one gets left too far behind while they’re deep in thought.