eggcorn

[Arnold M. Zwicky]

On May 19, 2005, at 8:35 AM, Rex W. Stocklin wrote:

> ... As I understand it, most grammarians widely accept "Ain't"
> now, as they do prepositional sentence caps. As well as split
> infinitives.

that depends on what you mean by "accept".  everyone who studies
english usage recognizes that "ain't" is very widespread, and has
been for a long time.  everybody also says that it's non-standard.

on the other hand, stranded prepositions and split infinitives have
been standard english (at least in certain contexts; in some
contexts, they are in fact obligatory) for centuries, and this fact
is recognized by everyone who studies formal standard written english
as it is actually produced by practiced writers.  the blanket "rules"
against these constructions are inventions without basis in fact.

> But I have yet to read or hear any groundswell take on
> acceptance of just any malaprop being accepted because of widespread
> misuse.

no one proposes such a thing.  many features of english are like
"ain't": very widespread, but stably non-standard for long periods of
time.  they are, however, none of them "mistakes" the way malaprops
(of either variety) are mistakes.  people who use these features
aren't aiming for something else and getting these instead.  they're
saying (or writing) exactly what they mean to.

on the other hand, a huge collection of language changes begin in
"mistakes" of one kind or another.  everything called "reanalysis"
and "grammaticalization", in particular, turns on some people
mistaking aspects of the gramars of  people around them.  such
innovations begin life as minority options (and may be disapproved
of), then catch on and spread rapidly, and eventually become part of
the standard language.

arnold, saying very elementary things about language