As part of the 7 Things Meme,
sandmantv asked me to write about Prescriptivism vs Descriptivism. But I really can't do any better than link to
today's Slate article by Steven Pinker, which is an excellent debunking of common misunderstandings of these two viewpoints. It's long, but I highly recommend reading the whole first section (approximately the first two pages).
I will add one addendum/clarification to the article, specifically following on this paragraph:
But the valid observation that there is nothing inherently wrong with ain’t should not be confused with the invalid inference that ain’t is one of the conventions of standard English. Dichotomizers have difficulty grasping this point, so I’ll repeat it with an analogy. In the United Kingdom, everyone drives on the left, and there is nothing sinister, gauche, or socialist about their choice. Nonetheless there is an excellent reason to encourage a person in the United States to drive on the right: That’s the way it’s done around here.
For the most part I like this analogy, but I am worried that the use of country-wide conventions as an example will give the mistaken impression that language variation must also occur at the national level. In fact, language conventions vary on a much smaller scale, such that many people switch between linguistic communities on a daily basis. The relevant scale here is "linguistic communities", which could be your family, your neighborhood, your state, etc.
We (and here I speak for myself, Pinker, and the general community of reasonable non-iptivists) are not arguing that communities of non-standard dialect speakers should switch to Standard American English. If a community speaks African American Vernacular English, for example, that's fine, and we are not arguing that they should all switch to SAE because "that's the way it's done around here [in the US, broadly speaking]". Rather, when a speaker is within that community, I fully expect them to speak in AAVE, because that's the way it's done around here [in the AAVE-speaking community].
However, for someone who grows up in the US primarily speaking a dialect other than SAE, it is also beneficial for them to learn SAE, because it's the standard in education, business, politics, etc. That doesn't mean they need to speak it at home or with their friends, but having the ability to use it (both in speaking and writing) is a crucial skill in our society.
There is huge untapped potential for teaching about language variation in early education. Discovering differences between people's language usage is fascinating, and it's a topic that's ripe for lots of in-classroom discussion and exploration. (Obviously I would think so, but I swear that my non-linguistic friends find it interesting as well.) The key is to treat all language variants on an equal footing, making it clear that no variant is inherently better or worse than any other. With this background, you can also teach the conventions of SAE, with the understanding that there is nothing inherently special about these particular conventions except that "that's the way it's done around here [in certain parts of society where you may want to get a job someday]".