Live data from Hacker News

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

gwern.net

261–270 of 404 posts

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

#261

Earlier quoted context omitted.

Makes sense to me. But my date and place of birth don't define me, they only correlate me with others. Astrology's just a particularly bogus way of not doing that, but anything that would try to derive anything about me from just my date and place of birth, without any actual data to correlate with my answers, isn't really worth my time. i.e. you'd need to know "people born around this date in this place tend to have…

> Astrology's just a particularly bogus way of not doing that, but anything that would try to derive anything about me from just my date and place of birth, without any actual data to correlate with my answers, isn't really worth my time. Well, it depends on the degree which we require it to define you. In a looser degree, it would define most people quite a lot. Put another way your "place of birth" alone, would be…

There must be some sort of misunderstanding here.

> If we were to bet on whether a person is black or white, freckled or not, has epicanthic fold or not, etc, I don't need no special "data collected" to know what people born in Lagos vs Dublin would look like for example.

Yes you do. In order to make an educated guess, you'd have to know, given where they were born, whether that place is more likely to produce certain traits or not. You could then use that information to make your guess about how likely they are to possess those certain traits, having been born there. But without any information on how likely those traits actually are for people born there, knowing where they were born grants you nothing.

> It's common knowledge.

It's not common knowledge what my Human Design type would be given only when and where I was born. That's the whole point I'm trying to make. Astrology or not, the entire idea of giving me an answer solely based on when and where I was born, will always be bogus.

Now, if they had an actual dataset of the Human Design types of people from all around the world, and wanted to correlate me with that, then maybe I'd be inclined to give it a try, just out of curiosity.

But I'm not even the slightest bit curious about bogus astrology (or any other types of divination). I'd be glad to take a personality test, for example, because they ask actually relevant questions. Just not this.

> And in some cases where this changes over time, the data of birth would also come in handy. E.g. in the tendency of a random London citizen to have say South Asian features in 2024 vs 1950.

I'm not arguing against correlating this with other data. The entire point is that one needs to correlate it with other data in order for my answer to be at all useful, and what I have a problem with is that correlation is not being done. The alignment of the planets alone is not going to reveal my personality type, and not even when combined with my location.

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

#262

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…

Look at how different we are from each other. Then consider that humans actually have less genetic variance than most mammal species. Now imagine meeting actual aliens whether from space or from our own AI efforts.

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

#263
In college I was struck by a housemate's law books. They only contained text with no pictures, drawings, diagrams, tables, graphs, formulae, or any of the other tools for organizing information common to engineering. They looked to be extremely tedious and boring.

Later, another former housemate, a math major who was developing software, went to law school for a JD. It was his opinion that many of the tools from computer science were useful for deciphering the more complex laws and regulations.

How you think may be a significant factor in whether you are a good match to certain occupations.

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

#264
From what I read in this paper, it seems like the authors are depending on some very strict definition of what can be considered language. Can anyone provide more context on what definition they're using?

I ask because certain assumptions that seem to be built into the paper and its references seem to exclude a lot of things that I would personally consider other forms of language. For example, they say the subjects with impairments to the parts of the brain supposedly required for speech could "follow non-verbal instructions" and "understand what another person believes". What makes those exchanges distinct from use of language, albeit a poorly defined one? I know nothing about this field of study, so I assume there's some assumptions and definitions they aren't stating explicitly. It seems weird to me that they say "these representations need not be specifically linguistic: they could be symbolic but non-linguistic (for example, ‘9’), and the use of symbolic non-linguistic representations does not engage linguistic resources (for example, mathematical reasoning elicits no response in the language brain areas and is preserved in individuals with severe aphasia". Why not go back and question the initial assumption that all language depends on those specific parts of the brain? Why are symbols not language?

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

#265
post #178

Earlier quoted context omitted.

So you're effectively saying that Chinese is compressed Tamarian ? Like, instead of writing "Darmok and Jalad at Tanagra" you'd put their initials together, so you're saying the Chinese-equivalent of "DJT" and the listener instantly knows the story, and therefore the message, it refers to?

You recognized it. Tamarian does exactly that. There's a subclass of 成语 the sense of which you can't even guess at unless you know the story: - 塞翁失马 (Sai Weng Shi Ma) - Old man loses horse; a 5-fold story of riches-to-rags-to-riches-to-rags-to-riches = life is unpredictable - 自相矛盾 (Zi Xiang Mao Dun) - Resemble spear and shield = unstoppable force meets indestructible object The latter is so common that 矛盾 (Mao Dun) i…

"Seven at One Blow" is certainly not a proverb that everyone knows in English; it's the name of the fairytale in English, but not something that people use to mean "bamboozle".

It's simply the case that Chinese relies far more on literary/historical allusions (chengyu) than English does. We might talk about someone's "Achilles heel" or a "Trojan horse", but these literary/historical allusions are simply nowhere near as common as chengyu in Chinese.

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

#266

Earlier quoted context omitted.

> it purportedly uses date and place of birth, but those hardly have anything to do with me While astrology is bogus, you'd be surprised how much "date and place of birth" has to do with you. There are statistics and studies showing higher than chance similar behavior/tendecies in people born in the same months (for things like depression, health outcomes, etc). Could have to do with exposure to sunlight during early…

An interesting test would be to correlate birth date results with people from the southern hemisphere. "Western civilization" is very north biased IMHO.

Human civilization is very north biased. Almost 90% of all people live in the northern hemisphere.

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

#267

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…

Disclaimer: I do not know, really, what role language plays in a thought process. I just want to point why your takes are not enough to make me to believe that language is a necessary prerequisite for a thought.

1. Rubber duck maybe just help with attention issues. All the studies I heard of do not try to untangle the mechanism. People tend to use language for a multi-brain thinking, and in this mode people do not think their thoughts fully, they propose ideas allowing other to support them or dismiss them. In this way they've got combined knowledge and experience to do the work, simplifying the early rejection of bad ideas. And I'm sure this mode of thinking shapes mind and in particular attention processes. You need to track which bits of information you told already and which you didn't, and you trained for that. Rubber duck can be just a trigger for that mode of attention.

2. I didn't hear about externalities, but to my mind what really helps to remember it is a number of associated details. I believe that ideas extracted from the memory when some of these details is popped up in your thought. It serves like a key in an associative map. When you name a concept with a word, and then use this word in different combinations with other words building associations, then sometimes your mind just like LLM will suggest the first word when you used words associated with it. It seems like externalities you talk about. Language plays its role with this, but you can achieve the same result without a language but thinking about all connected concepts of a concept you are trying to remember. You can build associations this way.

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

#268
post #154

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 reminds me of this old Feynman video where he talks about how people use different methods in their heads when counting, seems relevant: https://youtu.be/Cj4y0EUlU-Y?t=135

+ Rudolph Flesh, The Art Of Clear Thinking (1951). He has visualizations in there how people concepualize the months in a year as an ellipsis, and so on.

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

#269
post #24

Earlier quoted context omitted.

> * 关系 (Guanxi) (Mandarin): Your network of connections that help you get stuff done in life and business. I do think in American English at least there is a phrase for this: Your Network. You even used it to explain the word :) Usually the first association to “your/the network” is social. You have to specify if you mean the technical version with computers.

The mandarin guangxi encompasses a whole lot more than just your network. That would be considered a very bastardised translation, losing all the extra meanings such as good luck created by having that network, possibilities to fortune, a kind of karma, and more.

Karma aside, "having connections"/being "well-connected" is pretty much the equivalent in English. It's not bastardised at all; terms don't need to have the exact same cultural implications to be basically equivalent.

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

#270

In college I was struck by a housemate's law books. They only contained text with no pictures, drawings, diagrams, tables, graphs, formulae, or any of the other tools for organizing information common to engineering. They looked to be extremely tedious and boring. Later, another former housemate, a math major who was developing software, went to law school for a JD. It was his opinion that many of the tools from comp…

> the tools from computer science were useful for deciphering the more complex laws and regulations.

I've sometimes thought that certain aspects of law (e.g. contracts) are very similar to the strict rule-based thinking required in programming, and that the same basic logical thinking is required. It's just that some concepts have been formalised differently between the fields, and also that law has inherent fuzziness due to incomplete specification of its programs, the ambiguity of some language/phrasing, and the changing societal environment.

With programs, the computer interprets the rules deterministically (generally speaking), and we programmers are forced to draft our regulations accordingly or we'll get unwanted results. Whereas contracts are a bit less deterministically interpreted, and human incentives and potential interpretations have to be considered in drafting.

Post reply on HN