|
|
Post by Baron von Lotsov on Feb 29, 2024 15:54:33 GMT
We all know who Chomsky is and that is he is a lefty who invented the oxymoron Military Intelligence, which previously had been seen to make sense. He had a lot to say about the military industrial complex and manipulation of the masses. In fact even though I would not identify as a lefty myself, I do have a respect for his intelligence, and to me he never looked like a bad guy, like not prone to fibbing, as many political activists do.
Some may have read his biography and noted he was a linguist. I've known it for a long time in fact but this is where it stops. The bios say, oh he was a linguist as a kind of footnote. So what you may ask. Well I've been trying to study some maths recently. I can't say I have got my head around it properly, more like just dipped my toes in, but bits of it are starting to make sense so I push on. My reason for dong this is because I'm trying to invent a computer language. Now before you scoff and think oh yeah, well I've got as far as having something that does really work and can be used for useful applications, so in a sense I've already invented it. The thing is the system I have is limited and I'm working on extending its capabilities. To do this I've been trying to learn this maths so I have some tools upon which to engineer this language into its final form.
To find what you are looking for can be a bit of a treasure hunt since maths is a huge subject and I do not have my own linguistics teacher to point me here of there. So as I was looking at these strange things called monoids I start to see a place where formal maths and computer science meet. The maths is very much concerned with set theory and group theory plus their theoretical Turing machine which is used to help analyse this and that. So I was kind of homing in on the maths needed for the job and all of a sudden Chomsky turns up. Chomsky approaches it from a linguistic tradition, the mathematicians approach it from things like category theory, and at a point the one meets the other and sees they both got the same answer. As a result a lot of formal language theory was developed around this time leading to the development of the first computer languages, like Algol is often mentioned. There were some other funny ones as well like Lisp and small talk.
Anyway, what Chomsky did is now called the Chomsky hierarchy. Wiki is a bit "work in progress" on these pages, but this is how it is described.
You can trace through the links on this page to get a lot of the other theory, since this a top-level overview.
You'd never have thought maths and English have anything to do with one another, but they do.
|
|
|
|
Post by Vinny on Mar 4, 2024 7:53:02 GMT
That silly old fool defended Pol Pot and denied the Cambodian genocide.
|
|
|
|
Post by Baron von Lotsov on Mar 4, 2024 21:29:26 GMT
That silly old fool defended Pol Pot and denied the Cambodian genocide. He's definitely a fav of the socialists. All I really know about that time was Pol Pot was like a result of US war actions in Vietnam radicalising them, because Cambodia was being used as a supply route that the US were bombing. It has to be one of the worst regimes in history that one. |
|
|
|
Post by Bentley on Mar 4, 2024 21:41:41 GMT
I’ve read a couple of his books and listened to quite a few lectures . I even tried to get to grips with his spin on linguistics.
Very interesting man . He kind of spoils his credibility by saying stupid things such as ‘ Jordan Peterson views are right of Attila the Hun’. I think he is just too old and tired now.
|
|
|
|
Post by Ripley on Mar 4, 2024 21:49:46 GMT
For linguistics, I much prefer Steven Pinker. I find Chomsky a bit inaccessible, and possibly deliberately so. Pinker has written some excellent books.
|
|
|
|
Post by sheepy on Mar 4, 2024 21:59:20 GMT
I am more of the Beano type reader, but I will watch and learn from the intellectuals.
|
|
|
|
Post by Baron von Lotsov on Mar 5, 2024 4:57:52 GMT
For linguistics, I much prefer Steven Pinker. I find Chomsky a bit inaccessible, and possibly deliberately so. Pinker has written some excellent books. During the 60s and 70s there was a lot of research into artificial intelligence. They tried to do it with a deterministic computational method which in part should be able to understand natural language by encoding the rules of grammar. It looks like this is roughly where he was at in his line of research.
They were wildly overoptimistic in assessing the difficulty of the job, so by the 80s AI was seen as a pipe dream. The actual solution they needed all along is how we do it today, which mathematically is trivially simple compared to the rat's nest of complexity they had explored. The only problem now is the way we do it, although it works, it is not really understood how it works. These simple ideas we have now which work were eluding them all along because they had just gone down the wrong path, which is how Bertrand Russell felt after his Principia Mathematica. Anyway, it is finding other uses now, especially for computer language design.
|
|
|
|
Post by Ripley on Mar 9, 2024 15:33:17 GMT
For linguistics, I much prefer Steven Pinker. I find Chomsky a bit inaccessible, and possibly deliberately so. Pinker has written some excellent books. During the 60s and 70s there was a lot of research into artificial intelligence. They tried to do it with a deterministic computational method which in part should be able to understand natural language by encoding the rules of grammar. It looks like this is roughly where he was at in his line of research.
They were wildly overoptimistic in assessing the difficulty of the job, so by the 80s AI was seen as a pipe dream. The actual solution they needed all along is how we do it today, which mathematically is trivially simple compared to the rat's nest of complexity they had explored. The only problem now is the way we do it, although it works, it is not really understood how it works. These simple ideas we have now which work were eluding them all along because they had just gone down the wrong path, which is how Bertrand Russell felt after his Principia Mathematica. Anyway, it is finding other uses now, especially for computer language design.
That must be nearly impossible, because not all languages share a common family, and grammar rules vary from language to language. In Pinker's book The Language Instinct, he examines how children acquire language, and grammar in particular, by a mechanism that is not fully understood. Given a wholly artificial word, children can figure out how to use it depending on its context. But progress must have been made. I am going to be travelling soon and was given a language translator because I'll be visiting a country whose language I have not acquired. I've been playing with this gadget, which is the size of a mobile phone, and testing its accuracy against languages that I do know. To my surprise, it's quite accurate. Google Translate, on the other hand, is still abysmal. |
|
|
|
Post by Baron von Lotsov on Mar 9, 2024 18:05:38 GMT
During the 60s and 70s there was a lot of research into artificial intelligence. They tried to do it with a deterministic computational method which in part should be able to understand natural language by encoding the rules of grammar. It looks like this is roughly where he was at in his line of research.
They were wildly overoptimistic in assessing the difficulty of the job, so by the 80s AI was seen as a pipe dream. The actual solution they needed all along is how we do it today, which mathematically is trivially simple compared to the rat's nest of complexity they had explored. The only problem now is the way we do it, although it works, it is not really understood how it works. These simple ideas we have now which work were eluding them all along because they had just gone down the wrong path, which is how Bertrand Russell felt after his Principia Mathematica. Anyway, it is finding other uses now, especially for computer language design.
That must be nearly impossible, because not all languages share a common family, and grammar rules vary from language to language. In Pinker's book The Language Instinct, he examines how children acquire language, and grammar in particular, by a mechanism that is not fully understood. Given a wholly artificial word, children can figure out how to use it depending on its context. But progress must have been made. I am going to be travelling soon and was given a language translator because I'll be visiting a country whose language I have not acquired. I've been playing with this gadget, which is the size of a mobile phone, and testing its accuracy against languages that I do know. To my surprise, it's quite accurate. Google Translate, on the other hand, is still abysmal. It turned out to be far more difficult then they had imagined. It's the same with a similar thing they called expert systems. Encoding the rules by hand is just too complicated. This is the current approach, ironically for what you said, it was invented by Google.
|
|
|
|
Post by Ripley on Mar 9, 2024 19:35:30 GMT
That must be nearly impossible, because not all languages share a common family, and grammar rules vary from language to language. In Pinker's book The Language Instinct, he examines how children acquire language, and grammar in particular, by a mechanism that is not fully understood. Given a wholly artificial word, children can figure out how to use it depending on its context. But progress must have been made. I am going to be travelling soon and was given a language translator because I'll be visiting a country whose language I have not acquired. I've been playing with this gadget, which is the size of a mobile phone, and testing its accuracy against languages that I do know. To my surprise, it's quite accurate. Google Translate, on the other hand, is still abysmal. It turned out to be far more difficult then they had imagined. It's the same with a similar thing they called expert systems. Encoding the rules by hand is just too complicated. This is the current approach, ironically for what you said, it was invented by Google.
That's really interesting, and for someone like me, who has learned languages in the traditional way, a rather novel way of thinking about the building blocks of language. |
|
