Hacker News new | past | comments | ask | show | jobs | submit login
On Language and Humanity: In Conversation with Noam Chomsky (mit.edu)
91 points by anarbadalov 70 days ago | hide | past | web | favorite | 31 comments

It's weird to think that Noam Chomsky, who is so well known for his politics, and is reasonably well known for his contributions to linguistics, is also a significant contributor to automata theory & programming language design, despite having a very minimal direct involvement in CS. Imagine strongly influencing an entire field and not really working in that field. What an intellectual giant.

"Reasonably well-know" for linguistic theories.

Sure, if "most cited academic author of the 20th century" is reasonable [1].

Also, there's a deep connection between automata and abstract languages - an abstract language can be described by the class of automaton that can accept it and vice-versa, at least in the case of finite automaton, pushdown automaton and Turing machines go, you have regular, context-free and context-sensitive language (known as ... "the Chomsky Hierarchy").

[1] http://news.mit.edu/1992/citation-0415

Hi - so I totally agree with you, he's a titan of linguistics. I meant well known beyond the academy; I'd wager he's more of a household name because of his politics than his contributions to linguistics.

I heard about his all human languages look the same to a Martian antropologist at primary school.

For anyone who'd want to know more, this is the textbook I used in a CS theory class where I first learned about the Chomsky connection: Introduction to the Theory of Computation by Michael Sipser (https://www.amazon.com/dp/113318779X/ref=rdr_ext_tmb).

(Edit: added the title and author name in my post)

Sipser is a gem for anyone interested in formal languages. Also check out Harry Porter on YouTube for his great explanations: https://www.youtube.com/playlist?list=PLbtzT1TYeoMjNOGEiaRmm...

>reasonably well known for his contributions to linguistics

It's been 20 years since I finished my BS in Linguistics, but back then his name came up in just about every meeting of every class. We spent a lot of time discussing/arguing about whether he was right or wrong. I had one professor who was convinced that Chomsky was wrong, but he hired a TA who was a big fan of Chomsky so that we could get 2 different perspectives.

> but he hired a TA who was a big fan of Chomsky so that we could get 2 different perspectives.

That's... rare, and beautiful. I admire being open-minded like this. I guess you felt in a similar way?

He's probably spent the least amount of time dealing with politics but it's what grabs him the most attention. This also causes lots of people to dislike him for no good reason; there was a VC on twitter recently who basically said his work on linguistics and politics have been debunked and then proceeded block a few engineers who (rightly) pointed out his considerable contributions to PLT.

>He's probably spent the least amount of time dealing with politics

I'm not sure about that--he's written a lot of political books about particular countries/wars, each of them lengthy and painstakingly researched and footnoted--they alone would have been an ample life's work for most people. (I mean the earlier books, not the recent ones mostly based on interviews/conversations.) Then there's all the speeches, conversations, interviews etc on political topics.

His memory is remarkable. Even when he talks about politics he seems to know and remember more than virtually anyone else I’ve ever seen.

> This also causes lots of people to dislike him for no good reason

They're certainly not a "good" reason to dislike him, but the reasons are very real and significant.

In the case of the VC, Chomsky spreads ideas that would destroy their economic power and give it the the people. People with millions of dollars in wealth are terrified of anarcho-syndicalists like Chomsky.

It's fascinating that to get attention and acceptance for his theories, Chomsky had found modern linguistics in the concrete sense of creating departments, training students and so-forth.

Over the last few years, I've become more aware how dependent ideas are on fitting into institutional frameworks - some are outside such frameworks but without people to systematically develop them, such ideas tend to be ephemeral.

Will a better-connected, more searchable record of academic output allow talented thinkers to sidestep this historic need for academic track hoop-jumping and obtain reasonable assurance of future recognition for independent work?

No, because search isn't the problem, curation and education is. Search by itself is a waste of time, if you don't know what you don't know, and don't know what to look for.

Institutional frameworks solve those problems.

Indeed, telling a person the those thing "they don't know they don't know", telling people the key question people are trying to answer in the field, provide structure and guidance while a person spends time attempting to address a given question, etc, etc, etc.

All these things are powerful and shape most modern intellectual endeavors. And by that token, someone who apply discipline and rigor in approaching questions outside an institution can have an incredible effect - back to Chomsky essentially creating the field that he's dominated (theoretical linguistics, one thing I didn't know 'till this article).

The key question that people are trying to answer in the field of fluid dynamics is how to find the solution to the Navier-Stokes set of equations.

There you go. You have the question. You have a search engine. Go solve it. Oh... Hmm.

'Knowing which question to solve' only helps for easy problems. A search engine can help you solve easy problems. It won't do a damn thing for hard problems. You need someone to curate what is known in the field, point you in the right direction, steer you away from dead ends, help you break the problem down into bits that you might make headway in.

Academic institutions do just that. They don't always do it optimally, they can get stuck in local maximums, they are occasionally blind-sided by some genius that attacked the problem from an unorthodox angle.

But they do a hell of a lot more then a search engine ever would.

The whole notion that there is a field and it has a key question and there's a goal to solve it is already a triple set of local maximums.

Not sure I agree. There are plenty of curated resources and institutional frameworks but the last time I took an online university-affiliated class I would have learned more and saved time by simply reading a book. I think being broadly educated is partly having the capacity to delve fruitfully in to unknown domains and grapple with unknown semantics. Perhaps I am biased: my mother was a librarian.

To me, the most important work in linguistics in my lifetime has been the modern ability of computers to convert spoken language to written language and thus use it to control software, which I use daily in things like Alexa and CarPlay driving directions. What I’m not sure about is, has the Chomskyan style of analyzing language been useful for that at all?

Basically, a lot of pre-1980 theories of mind seemed promising at the time because they had some explanatory power and made sense. But more recently, machine learning techniques of treating the mind more like a large-dimensional statistical object rather than something with complex internal structure seem to be proving more effective. I wonder whether in time the Chomskyan style of linguistics will prove to be a dead end.

I just have to say that, assuming this was a spoken interview, Chomsky can talk. I can scarcely comprehend using the same number of choice phrases Chomsky employs in a sentence if I had a paragraph - and I certainly wouldn't be able to do it at the speed of speech.

I've heard of people disgruntled by his economics and political activism denigrating him as "not particularly smart". This article is proof to the contrary, politics notwithstanding.

he keeps blowing me away. even at 90. seriously...who can come remotely close to his breadth, dept, and consiousness?

i am sad to see him go sooner or later, and also sad that the left has no replacement for him.

Interesting how he bashes ML translation as having little scientific interest

Anybody who speaks more than one language knows to not trust translations(even those made by humans), especially if they're made by some black box algorithms. It's one of those tasks where quality is almost (objectively) immeasurable and perfection is impossible.

Is it just me, or whenever I try to read what Noam Chomsky says or listen to him I have to constantly read over it to understand it?

I don't know if it's true for Chomsky, but I have noticed that a lot of academic sentences use a lot of inner clauses. I think it's a bad habit. When reading those sentences as a programmer, I wish for a lot of parentheses so I can determine order of operations. Without them we're left to infer that order, which is extra work when we're also just trying to figure out what the sentence itself is about.

Here's an example from T.M. Scanlon's "What We Owe to Each Other", a philosophical book on morality I was reading yesterday:

"When I ask myself what reason the fact that an action would be wrong provides me with not to do it, my answer is that such an action would be one that I could not justify to others on grounds I could expect them to accept."

This sentence is actually key to what the entire book is about, but.... "what reason the fact that an action"? I mean even as programmers we are taught to break apart long boolean expressions.

It's kind of like garden path sentences but a more general syntactic ambiguity that smart people are more guilty of due to their need to explain complex subjects. Hard for a writer to recognize, because they know what they mean.

I would go further and say that many academics hide their banal or confused ideas behind overcomplex language. They learn to write that way by reading all those awfully written texts and imitating them. But Chomsky is not like that - he usually has something insightful to say expresses it plainly.

... AND expresses it plainly. Oops.


Given that Chomsky is an extraordinarily accomplished intellectual, I think it's a little denigrating to assume if you can't easily understand his works you read only children's book.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact