Live data from Hacker News

Language is primarily a tool for communication rather than thought [pdf]

gwern.net

331–340 of 404 posts

Re: Language is primarily a tool for communication rather than thought [pdf]

#331

My takes: 1. The existence of “rubber duck debugging”, and a whole bunch of studies on verbally explaining a concept indicate that language is essential for thought. In rubber duck debugging, programmers tell their problem to an object and this is beneficial in finding the solution. There are studies that show when we verbally teach material to someone we remember it better. Also the act of taking a test increases le…

These were my exact thoughts when i read the headline. But when i read the article, their strongest evidence in their defense is the fact that people with severe disabilities can still perform higher level thought.

tldr; Language is not required for thought, but on page 3 of the link, "Instead, these tasks engage other brain areas that are non-overlapping with the language network (Fig. 1b), although they sometimes lie in close proximity to the language areas"

I used to do rubber duck debugging, but instead i turned into listing exact step-by-step instructions in inline comments. It seems to work like rubber duck debugging, and all it took was for me to "Externalize the steps" so i can think about what needs to happen at each level of the thought heirarchy.

and section 2. you describe memory and thought as similiar in nature but those are handled in two entirely different regions of the brain.

Re: Language is primarily a tool for communication rather than thought [pdf]

#332
IIUC this is in direct contrast to a previous thread: https://news.ycombinator.com/item?id=40466814

That thread argued that before hellen keller was taught language in some form she was not conscious. Even the top rated comment seems to corroborate it.

So which one is true?

Re: Language is primarily a tool for communication rather than thought [pdf]

#333

It absolutely melts my mind every time I come across the two facts that: - People experience their thoughts very differently - We all secretly believe that deep down, everyone experiences thought like we do. I've never really had a strong internal monologue when thinking, so my assumption would always be that of course, thinking isn't very linguistic (even if we can use it as a tool while thinking). It seems like the…

This is a big part of why I think LLMs can lead to AGI: humans clearly exist who can think equally well despite having some major gears in their brain going in different directions from each other. If you told me that there was a human who could only think by speaking and who experienced their entire reality and cognition in a textual form, occasionally augmented by flashes of visual or auditory data, I would not be willing to comfortably put an upper limit on their IQ despite all that. I'd just say "yeah, I guess? Weirder things have happened."

Re: Language is primarily a tool for communication rather than thought [pdf]

#334

Earlier quoted context omitted.

> Another small number of people talk about imagining complete objects out of nowhere that they then have to analyse to describe; the imaginary thing has complete shape before its shape is described. But this is different from the kids tea party. I can easily pretend to hold a teacup without closing my eyes and visualizing a teacup, I don't think one has anything to do with the other. If you ask someone to actually i…

If I cover your eyes and put a teacup on your open palm so you can't see the colour or feel the size, _what are you holding_? I'm imagining a teacup - made of china, decorated in some way, capable of holding a few mouthfulls of ~75C thin liquid, smaller than a big mug, larger than a shot glass, with a handle sizes for the tip of one finger, delicate, ceremonial; concept space constrained to broadly teacup-ish area. B…

> If I cover your eyes and put a teacup on your open palm so you can't see the colour or feel the size, _what are you holding_?

I obviously failed to communicate what I was saying.

If you give me a teacup and tell me to imagine I can see it, I have to know what it looks like before I can form that image.

This is all just a failure of communication because to me, part of the definition of having an image of something in your mind is that you know what it looks like.

Since knowing the colours and shapes are required to form an image, if you don't have those, you don't have an image.

Imagining a teacup is possible without imagining you can see a teacup.

Re: Language is primarily a tool for communication rather than thought [pdf]

#335

Earlier quoted context omitted.

> Another small number of people talk about imagining complete objects out of nowhere that they then have to analyse to describe; the imaginary thing has complete shape before its shape is described. But this is different from the kids tea party. I can easily pretend to hold a teacup without closing my eyes and visualizing a teacup, I don't think one has anything to do with the other. If you ask someone to actually i…

> If you ask someone to actually imagine seeing a teacup but what they're imagining doesn't have a colour or size, _what are they imagining_? The pure, abstract essence of teacup, to them. I suppose I assume it does have size and shape (it's functionally an imaginary cup, after all, and not another imaginary object). Maybe to someone who has only ever seen blue teacups, it's inherently blue? For me, if I start that e…

My point is this: if I tell you to close your eyes and form the image of a teacup in your mind, images certainly have colour and shape, so either your image does too, or you don't have an image. It's really just down to a differing definition of image.

If you ask me to imagine seeing a maze with hundreds of turns, rather than saying I have a nonspecific image, I'll say I'm having trouble forming a concrete image.

If you ask people to imagine a teacup, some will imagine seeing one, others will imagine holding one, others will have no images or feelings in their mind.

I'm pretty sure, like others are, a lot of this is just down to ambiguous or poorly specifies language.

Re: Language is primarily a tool for communication rather than thought [pdf]

#336

Earlier quoted context omitted.

> If you ask someone to actually imagine seeing a teacup but what they're imagining doesn't have a colour or size, _what are they imagining_? The pure, abstract essence of teacup, to them. I suppose I assume it does have size and shape (it's functionally an imaginary cup, after all, and not another imaginary object). Maybe to someone who has only ever seen blue teacups, it's inherently blue? For me, if I start that e…

My point is this: if I tell you to close your eyes and form the image of a teacup in your mind, images certainly have colour and shape, so either your image does too, or you don't have an image. It's really just down to a differing definition of image. If you ask me to imagine seeing a maze with hundreds of turns, rather than saying I have a nonspecific image, I'll say I'm having trouble forming a concrete image. If…

> My point is this: if I tell you to close your eyes and form the image of a teacup in your mind, images certainly have colour and shape, so either your image does too, or you don't have an image. It's really just down to a differing definition of image.

Not really. I mean, I have some aphantasia-type aspects with faces. But my mind's eye is quite amenable to the semi-lucid mental image of a cup that doesn't have colour until I ask myself what colour it has, doesn't have fine shape details until I consider them.

After all, not all of our eye even sees in colour! Part of our visual apparatus only provides black and white imagery to our brain. So there's plenty of colourless imagery within our cognition.

Re: Language is primarily a tool for communication rather than thought [pdf]

#337

In my limited (to one person) experience, it may not be as clear cut. I am one of those who uses language to think. I would have an internal monologue and I would also say things out loud just to see how they sound. It's not always a fully fledged idea, or even a coherent thought for that matter. I have no idea why i do this. But that is me. In a sense, my thoughts are forming when I speak. Hence, i am somewhat bette…

I have similar problem. I don't have inner monologue, so when I want to share my thoughts, I try to compile them in real time during the speaking. Which means I have no idea how the sentence will end or if I will even be able to get to the point I want to make. And additionally I have problem with focusing (ADHD probably) so my only sensible way to communicate with people is by writing, not orally.

Re: Language is primarily a tool for communication rather than thought [pdf]

#338

Language is a major component of thought. Every programmer knows the unreasonable effectiveness of rubber duck debugging whereby one can solve logical problems more effectively by talking aloud. If language were merely a tool for communicating thoughts rather than an integral part of thinking, then talking aloud wouldn’t help. But since raising the volume of one’s internal monologue does improve reasoning, this shows…

Yeah, I am a programmer without inner monologue so it's not that simple.

Re: Language is primarily a tool for communication rather than thought [pdf]

#339

It absolutely melts my mind every time I come across the two facts that: - People experience their thoughts very differently - We all secretly believe that deep down, everyone experiences thought like we do. I've never really had a strong internal monologue when thinking, so my assumption would always be that of course, thinking isn't very linguistic (even if we can use it as a tool while thinking). It seems like the…

It's difficult to say what thought is, but LLMs also solve more problems more correctly if you allow them to feed back through themselves multiple times (ask them to show their working) apparently for the obvious reason that it allows more computation to happen. I hypothesize that different people build different recurrence circuits, by chance. Some go through the audio cortex, some through the visual. Helen Keller described herself as driven solely by emotions, like a beast, before she had any kind of language: https://www.reddit.com/r/woahdude/comments/1jyo53/hellen_kel...

Re: Language is primarily a tool for communication rather than thought [pdf]

#340

Earlier quoted context omitted.

Helen Keller wrote about having memories of being something like a stimulus-response automata before language enabled her to be conscious and think. I think that example is actually in favour of language being required for thought, because she remembers a state of being before language where she wasn't really conscious and didn't really think. I don't see any particular reason why language would be required for memor…

Do you remember where you read that? I'd love to read more.

Sure thing, here you go: https://scentofdawn.blogspot.com/2011/07/before-soul-dawn-he...
Post reply on HN