A few small but related questions here. I'm looking at ways to define "sentence patterns", at least starting with English. That led me back to phrase structure grammars, which have nice and fairly intuitive parse trees.
Main question that I've been wondering about is if all possible sentences can have an intuitive parse tree. I get that sentences can have many parse trees depending on context and the way you look at the sentence. But even so, is it possible to always represent a sentence using this system?
Finally, where do you learn all the rules for the symbols used in parse trees, and what word is considered what part of speech, sort of thing? I would like to more deeply explore complex sentence structures using this phrase structure grammar paradigm. To answer the question in my head about there maybe not being a valid parse tree for some sentences, or at least one that makes intuitive sense.
(Tangent question, wondering if there are open source projects which generate the data model for these parse trees given a sentence, I have spent an hour looking but will look some more).
For clarity, I'm thinking what it would be like if you wrote down all the possible parse trees. Like for "mary had a little lamb":
S > NP VP
NP > N
N > Mary
VP > V NP
V > had
NP > DET NP
DET > a
NP > AdjP N
AdjP > ADJ
ADJ > little
N > lamb
But wondering what can go into what, and generally the details of what's possible. So that's where I'm going for now.
