As the other linguistics geeks have clarified, there are two types of grammar: descriptive and prescriptive.
All languages have descriptive grammar. There is no such thing as a grammarless language - even languages of very “primitive” societies have grammars of comparable complexity to English. Even grammarless pidgins evolve a coherent grammar as soon as the first generation to be born in the pidgin society picks it up, which lends support to the universal grammar theory.
Some languages, especially literate languages such as English, have a prescriptive grammar. This is a wholly social phenomenon with no objective basis. Repeat: prescriptive grammar has as much to do with linguistic science as table manners have to do with gastroenterology.
However, it is evident that people who are not competent in the “standard” dialect are disadvantaged sociologically. For that reason, the grammar of the standard dialect ought to be taught in school.
(Note that in sociolinguistics, the term “acrolect” is used in preference to “standard dialect”; it simply means the dialect/sociolect that is privileged, without lending any particular legitimacy to it.)
I like the method that SpoilerVirgin described, in which children who speak a different dialect are taught the acrolect without having their own dialect denigrated; i.e. as a separate dialect which is useful to know.
Everybody speaks a dialect; it’s important to recognize that the acrolect has no objective legitimacy, just social.
IANA Expert-on-Black-English-Vernacular but I’d observe the following:
English singular and plural usage varies widely between commonly-accepted dialect. I presume, if you’re on this side of the Pond, you would not say “The committee are unable to reach agreement” as any British speaker with a knowledge of proper usage would be certain to say. Nor would you consider “two pair of pants” to be incorrect, though you may prefer “pairs” – and in any case would consider that this has reference to two distinct garments, not four, as “two pair(s)” in a poker game would imply four cards.
My distinct impression is that nouns of quantity do not automatically take the plural in BEV – 15 Lincoln-head coins would be “fifteen cents” (15 distinct coins) if they didn’t use “pennies” or “cent-pieces” to describe them, but a dime and a nickel would be “fifteen cent” (coins to the value of $0.15). What he might put in a college term paper has nothing to do with his vernacular usage.
More importantly, within the context of his speech pattern, the kid giving you “fifteen cent change” would be correct, and the kid who gives you “fifteen cents” is leaning over backwards to be extra-polite to a customer he is aware does not share his vernacular – and probably deserves thanks for the extra effort to be courteous to you, the alien in his world.
I’m afraid not. I go to high school right now and I learned how to diagram a sentence in freshman year. I think the value of learning how to diagram a sentence is limited at best. I don’t think drawing those diagrams made me a better writer. I may be able to identify nouns, verbs, direct objects, etc. but that is not a skill I think would require a picture to learn. It just makes things a little longer and marginally easier.
Just to toss a couple reference-type things in here…try either The Language Instinct and/or Words and Rules by Steven Pinker. They both have discussions about innate grammar and learned grammar. Plus, they’re pretty good reads.
And I’ll recommend Derek Bickerton’s Language And Species (University of Chicago Press). It addresses a point touched by matt_mcl, Lamia, and SpoilerVirgin – that while the form grammar originally takes in a language appears to be arbitrarily set (he cites studies of pidgens and creoles), grammar itself is intregal and essential to language. [sub]maybe even to consciousness, but that’s a whole other kettle of fish[/sub]
Bickerton also cites studies of language development in children, and their early struggles with syntax and plural forms etc. He suggests that these things show the human brain contains a number of templates for grammar, and that early clumsy (to our ears) language construction is no more than the trial and error process involved in finding the best template for the job.