[SPAM:#] learning "rules"

[Arnold M. Zwicky]

a brief report on a small project done by a student in my sophomore
seminar (on prescriptivism and usage) spring quarter...

the aim of the paper was to look at the effect of explicit teaching on
the way people perceive writing.  the student invented an *incorrect*
"rule" of english grammar -- "each" and "every" are grammatically
plural, and so require a plural verb -- and investigated how people
(stanford undergraduates, in this case, mostly freshmen) would give
explicit judgments of "grammatical
correctness" for sentences involving these two determiners.  cutting
away some details, there was a control group (who were not given the
"rule") and an experimental group (who were), and everybody took a very
brief pre-test questionnaire designed to check, among other things, the
subject's place on an autonomy/compliance scale.

the autonomy/compliance part was less interesting than it might have
been, since almost all of the subjects viewed themselves as
independent.

my student expected the subjects who were given the "rule" to tend
towards compliance to it; years of grammar instruction, often involving
very arbitrary-seeming rules, would have trained them to do their best
to conform to them, at least on tests.

the results were just the opposite.  the subjects who'd been given the
"rule" tended to be deeply suspicious of it -- in a post-test, some
explicitly rejected it -- and to fall back on their judgment as
speakers of english.  the control group, confronted by a mixture of
bizarre and unremarkable sentences, seem to have been made unsure of
themselves.

now, this is a small pilot study, with lots of uncontrolled variables
in it, but it's still intriguing.  one possible interpretation is that
explicit grammatical instruction, especially on points that don't seem
natural for the students (and perhaps especially for highly educated
people), fosters not conformity to rule but resistance.  if so, then if
you want to teach the conventions of formal standard written english,
you should carefully select the "rules" you explicitly articulate,
reinforce, and test: dwelling on split infinitives, stranded
prepositions, possessive antecedents for pronouns, "dangling modifiers"
that are in fact entirely natural, and so on might well  be precisely
counterproductive.

arnold (zwicky at csli.stanford.edu)