I would be interested in the people who did structural/functional diagramming and views to the concrete/solidity of "that's not proper english" or "that isn't how it works" because the other side of the coin is that english (and obviously other languages: Spanish and gender..) change over time, and are fluid against the needs of their speakers.
I guess I'm arguing that if you did training in the formalism of a parse-tree, I wonder if it tends to re-inforce a view in "proper" use of a language rather than it's emergent behaviour and shifts of meaning.
There are subtle parts of how English works that confuse a lot of people, that diagramming might help with. The most obvious to me was the Harry Potter scene where, in the movie, Professor Snape says "and even put a stopper in death" [0]. The book version was just "even stopper death".
These mean opposite things. One uses "stopper" as a noun, saying to prevent death. The other is using it as a verb, saying to bottle up death - create a potion to cause death.
I don't think this applies, particularly. Diagramming wasn't so much "this is how a sentence should be set up" as "this is the function of this word in this sentence."
No, that's a separate discussion. You can use the diagrams for both description and prescription.
> What's your point -- some things can't be diagrammed?
Well, at most that some things can't be diagrammed naively, and perhaps need a more complicated diagram language.
> The first one is deliberately "wrong" for advertising effect, [...]
No, it's perfectly fine. It's called 'middle voice' and is a normal, if relatively rare, feature of English. Your Winston-ad example also seems like your correction makes it less grammatical.
While I'm not a prescriptivist, I think diagramming (and the rest of the strict English grammar education I was given in the 70's and 80's) helped me to understand why grammar rules exist and how they bring clarity to communication. I think that also helped me when learning to write code.
I guess I'm arguing that if you did training in the formalism of a parse-tree, I wonder if it tends to re-inforce a view in "proper" use of a language rather than it's emergent behaviour and shifts of meaning.