[some snipping]
I've been mulling over my unease with the categorisation, and -- rather
to my surprise, as I'd not realised this -- find myself concluding that
categories which separate usage errors on the basis of what the user
intended, rather than what the user actually wound up doing, strike me
as weak.
[quoted material snipped]
My problem -- which I don't wish to force upon others who disagree
(hence the attempt at polite dismissiveneess....) is that I don't think
that the fact that people are aiming for something, and have a sense of
the the meaning of that something, is a particularly valid basis upon
which to categorise errors.
I have the same problem with malapropisms as a category. As I
understand it, in order to qualify as such it's not enough for, say,
"elocution" to be replaced with "elucidation": the speaker has to be
attempting (and failing) to be grandiloquent, and is required to be
oblivious of the comic effect which results. That is, if the
substitution doesn't have that intent and effect -- if it was just a
slip, or an attempt at humour, or some sort of dyslexic function -- it's
not a malapropism, it's something else.
As I've been understanding the term, malapropisms are a fairly
unproblematic category: "wrong word", a production error. The main
distinction is the one between advertent and inadvertent ones: the latter
are slips which speakers or writers would and do immediately correct if
they notice the substitution. I don't think grandiloquence is required,
though it obviously lowers one's tendency to doubt and double-check a word
choice. Puns and wordplay aren't errors; they add a whole new dimension of
meaning to a phrasing. Spelling influenced by dyslexia would need to be
parcelled out, too. Eggcorns are then a subspecies of malapropisms.
But the outcome -- the replacement of "elocution" with "elucidation" is
precisely the same. As with eggcorns, the only difference lies in the
intent of the person making the error, and the classifier thus has to
make some judgement about what that intent was, and whether the error
qualifies as type (a) or type (b).
For me "elocution" for "elucidation" is a malapropism (if it wasn't
inserted by the spell-checker, that is). Full stop.
I'm unsure why you insist on "intent". I try to minimise thinking in terms
of intent and instead to rely on semantics: there's nothing in the
semantics of "elocution" that would warrant putting it in a slot where
"elucidation" fits. It's not a substitution that makes sense. And the two
words are vaguely similar, but phonetically quite a bit apart. Those two
points prevent the substitution from being an eggcorn candidate.
It's that principle of category division which strikes me as weak. When
faced with "just desserts" -- and judging that it's sometimes a typo and
other times an eggcorn -- my reaction is that whilst the reason for the
error is of passing interest, it would be better to base analytical
error categories on *what* error was occurring (rather than *why* the
error occurs).
That's an interesting point. But I find myself being quite unable to
perceive a clear line between the "why" and the "what", or at least the
"why" as far as it has a bearing on the "what". Then there's the "how",
the "to what end" etc.
Some time ago, I burnt my foot with hot coffee. My friends didn't consider
"I burnt it with hot coffee" a sufficient reply to the question "What
happened to your foot?"
If you describe something, you might phrase the description as facts
("what"), but those facts may relate to the context in which this
something occurs. The leap to causal or final explanations isn't
unproblematic. But again, I don't think "intent" is a crucial category in
defining eggcorns at all. Except if you classify all semantics under
"intent" somehow.
But surely the degree to which the definitions of a category minimise
the uncertainty of the grey areas is a reasonable test of the validity
(or at least strength) of the categorising principles.
Yes, but not minimise above and beyond everything else. Even those in the
hard sciences have parasitic data, noise and the like, that has to be
accounted for.
Natural language is messy. Is there any kind of language data that isn't
potentially distorted by events that are random or beyond the researcher's
control.
An analogy would be the division of urban settlements into "cities",
"towns", "villages", and "hamlets" -- divisions which lead to disputes
about relevant attributes. (Does a "city" need a charter, a cathedral,
a given population, a certain urban mind-set, all of the above, or none
of the above?)
For that reason, whilst one finds these terms being used colloquially,
one doesn't encounter them as a category set for urban analysis. In
analysing settlements, the breakdown would be based on measurable
factors -- population, facilities, infrastructure, etc. Trying to apply
the moniker "city" to one and "town" to another isn't a particularly
useful exercise for urban analysis, and so these are not used as
category definers.
How pertinent is this analogy for telling the eggcorns from the acorns,
chestnuts and other slippery phenomena? The problem with "city" seems to
be that the colloquial meaning is too vague in some areas where English is
spoken, and that where it is well-defined, different areas have different
definitions.
With eggcorns, we know reasonably well what they are, though the term has
been in use for a considerable shorter time than "city" and is a
description of a phenomenon that occurs naturally, so it is permissible to
refine the definition to include related stuff that hadn't occurred to
anyone before. Cities also don't occur in multiple examples. So the
question of whether a particular production illustrates an eggcorn isn't
analogous to the question whether, say, Manchester is a city.
Back to "intent": I may, in an informal way refer to someone "meaning to
say" this or that, but even an urban analyst might use the term "city"
once in a while, if the context and the limits are clear.
And as it is for urban analysis, so it is for error analysis. Whether
something is a malapropism, an eggcorn, or poor spelling doesn't strike
me as a particularly fruitful distinction if (a) the end result is the
same, and (b) one has no means of telling the difference.
We seem to be in disagreement about how much can be inferred from examples
and other available material. When someone writes "I hate what this guy
did to me -- I hope he'll eat his just desserts one day", we know at least
that the writer thinks of the "desserts" in the idiom as something edible.
People choose metaphors. Sometimes they insert homophonous but
historically unwarranted metaphors into existing idioms. I happen to find
this and related phenomena interesting. The messiness of the real-life
data is another matter entirely.
[remainder snipped]
Chris Waigl
--
blog: http://serendipity.lascribe.net/
eggcorns: http://eggcorns.lascribe.net/