What does Chomsky mean by "let the rules run freely"?

Question

In this YouTube video:

World Science Festival, “Mind Your Language: Thought, Metaphor and Imagination”. (September 30th 2021.)

The link is directly to the timestamp of the video where he says the following quote:

We use [language] for communication, of course, but it's basically an instrument of thought. [This is] a long tradition - I mentioned Galileo, but many others just assumed that language is what was sometimes called 'audible thought': it's the instrument of thought.

Now, if you think about our own language in use… almost all we do is language; construct thoughts in our mind. Sometimes we externalize them; sometimes we communicate, but that's a kind of a side-property of language.

Now we can pin that down much more precisely. When you look at the actual design and structure of language, what you find systematically is that when there are instances where there are conflicts between communicative efficiency and computational efficiency, computational efficiency always wins.

There are very simple cases like simple structural ambiguities. If I say 'flying planes can be dangerous', that’s structurally ambiguous. [Does that] mean the act of flying or the planes that fly?

Well, the point is that if you just let the rules run freely, computational efficiency [wins]… but they harm communication because [it] means [we] can't understand what somebody's saying. Do they mean to fly planes or planes that fly?

And there are many things like this. Some of them pretty serious which impede communication, but mother nature didn't care.

When evolution moves along, some change takes place [and] you get the simplest solution; simplest most elegant solution. [It] may happen to be dysfunctional, [but that's] not mother nature's problem…

That's the way - it seems - [that] language develop[ed].

What does the boldface-emphasized excerpt mean?

Answer 1

Cool quote - but the simple answer is pretty short.

He means the rules which govern how language works, where it comes from, or as he might be inclined to call it, the “language faculty” or possibly “universal grammar”.

More broadly, he is talking about his belief that one of the biggest explanative factors for the observed or emergent properties of language is computational efficiency - that since biological evolution is (purportedly) a ‘tuning’ system finding optima of some given design, Chomsky believes that evolution determines the optima for language features regarding communicative efficacy vs. energy, effort or efficiency.

The example he provides is (maybe) that there is no clear reason why, given how systematically structured language is, language did not also include a grammar structure to help resolve the ambiguity in the “planes fly” example sentence - by which he means to demonstrate, “our language could accommodate that, but it didn’t evolve that way because the opportunity cost (of having it) wasn’t worth it” - it would require a more complicated grammar structure, that we don’t need, since while ambiguities do lead to communication issues in daily life, perhaps it was not optimally adaptive to try to rectify them all - like a person with OCD trying too hard to organize everything (coming from a person with OCD).

I can’t vouch for or against the veracity of Chomsky’s claim, but I am reminded of what I read a pretty famous biologist say once, that the theory of human evolution is less of a scientific field than people may realize, under a particular definition of “scientific” - that it’s not something we can actually observe and reproduce on demand - so it is in some ways more of the domain of the freer thought of the humanities, even if projected outwards from observable evidence or other testable claims. I personally think studying evolutionary systems can be approached from a good angle using computer modeling, game theory and genetic algorithms, to try to “recreate conditions” under which human language-like structures or objects can emerge in the stochastic state space of a system.