Should we teach grammar at all?

Well, that's quite a question to ask now that I'm getting close to the end of my project, but it's a question that has been haunting me ever since I started my research. Does formalized grammar instruction really improve writing? According to Patrick Hartwell's February 1985 article in College English called "Grammar, Grammars, and the Teaching of Grammar," the answer is NO.

Hartwell wrote this in 1985, responding to years of research reports and essays that explore this question. The question, as Hartwell puts it, is this: "What does experimental research tell us about the value of teahing formal grammar?" (106). The problem, though, is that over the 75 years this question has been asked, this "experimental research has for all practical purposes told us nothing" (106). It would seem that research is done and conclusions are drawn, but then that same research is interpreted by others who draw different conclusions. Therefore, research doesn't solve the debate (107).

Hartwell then goes on to describe the different types of "grammar" that are taught:
1. Patterns of language that convey meaning
2. Description and anaysis of formal patters (linguistic science)
3. Usage
4. What we learn in school
5. Stylistic grammar (like the Strunk and White version) (109-110)

Grammar 1 is the grammar that we all, as native English speakers, have in our heads already. We know how things should be in the sentence, but we're don't know the specific rule for it (111). If we learn Grammar 2, it does not impact Grammar 1. Grammar 1 is quite abstract and can be difficult to explain. We can try to come up with rules to explain Grammar 1 (what is, actually, Grammar 4), but as we know, the rules that we teach our students often have one, if not many, exceptions (112).

An interesting study done in 1967 by Arthur S. Reber showed that "mere exposure to grammatical sentences produced tacit learning: subjects who copied several grammatical sentences performed far above chance in judging the grammaticality of other letter strings" (117). Hartwell also notes that people performed much worse when given specific rules to follow, like flow charts, to make decisions about what was "correct" in their writing (117). On the other hand, Herbert W. Seliger, another researcher, suggests that the rules themselves aren't all that useful, but when people think they are and internalize them, it allows people "to access the internal rules they actually use" (119). The reality is that the rules we are taught in school (Grammar 4), are COIK--Clear Only If Known; basically, if you already know the rule, the rule will make sense. If not, the rule is useless (119). It's also noteworthy that an application of a compicated set of rules is NOT what a maure writer does and relies far too much on rote memorization (120).

So, what are we to do? First, students tend to correct their errors when they read their work aloud (121). Second, "A broad range of cross-cultural studies suggest that metalinguistic awareness is a defining feature of print literacy" (122); or, we should get our students thinking about their thinking about language. This means that we should talk about the different types of literacies that are used for a variety of purposes in our culture (123).

Stylistic grammar (Grammar 5) allows us to teach about language's use in a variety of contexts as well as the "active manipulation of language with conscious attention to surface form" (125). When we approach the teaching of Grammar 5, it's all about playing with the language.

Ultimately, Hartwell notes, "The thrust of current research and theory is to take power from the teacher and give that power to the learner" (127). If that's true, then it would make sense to have students play with sentences and to direct that play so that the students internalize the rules through this play rather than giving them a list of rules to memorize. Should we teach grammar at all? Sure, but maybe not in the traditional way that was presented to us in school.