Fwd: "sort of" is elitist? (now with data)

[Arnold M. Zwicky]

and now a reply from paul kay at berkeley.  again, some comments from
me at the end.

Begin forwarded message:

> From: Paul Kay <kay at cogsci.berkeley.edu>
> Date: October 6, 2004 11:55:50 AM PDT
> To: "Arnold M. Zwicky" <zwicky at csli.stanford.edu>
> Subject: Re: Fwd: "sort of" is elitist? (now with data)
>
> ...  I'm afraid I don't know anything on the social distribution of
> "sorta" or "kinda".  But that doesn't mean such info doesn't exist,
> because back in 1984 when I wrote about kinda/sorta I wasn't
> interested in the social distribution issue and I haven't been reading
> sociolinguistic stuff for a long time. I wouldn't be surprised if
> actual usage data showed the ratio of "sorta" to "kinda" varied
> positively with, say, education. But I wouldn't be surprised if no
> such correlation could be found, either.
>
> But suppose there were such a correlation and the specific usages of
> Kerry and Bush showed the former to have a higher sorta/kinda ratio.
> Since Bush and Kerry both benefited from 24 karat educations, would
> this observation prove Kerry to be an elitist or Bush to be a
> hypocrite?  Kerry would be using his natural vernacular.  Bush would
> be affecting a fake working class authenticity to which he was
> entitled by neither birth nor experience...

-----

that is, suppose there is a social concomitant to "sort of" vs. "kind
of", and "sort of" is the "higher" variant, and there is an actual
bush/kerry difference, and it has bush with more "kind of" and kerry
with more "sort of".  this is a hell of a lot of supposing, but we
still don't have an interpretation of the (very hypothetical) data: the
difference could be that bush and kerry both deviate from the levels of
usage you'd expect given other social facts about them, or that bush is
at roughly the expected level while kerry deviates on the "up" side, or
that kerry is at roughly the expected level which bush deviates on the
"down" side.  interpretive choices like this crop up all the time, for
instance in the analysis of language and gender.  note that choosing an
interpretation depends on estimating the expected levels of usage.

two more caveats, as if there weren't enough already.  the first is one
i've already mentioned, namely the possibility that different uses of
"sort/kind of" have different social concomitants.  i think this is
likely, but on slim evidence.  i *believe* that (for me and some
others) the loose adverbial use of "sort/kind of" (as in "They're my
friends, sort/kind of"), which is distinctly colloquial and informal,
favors "sort" pretty considerably over "kind".  there might be other
data of this sort, which run counter to the hypothesis that "sort" is
the higher variant across the board.

the other caveat is that social differences often interact with
contextual differences, so that social values associated with
linguistic choices may be quite different, even opposite, in different
contexts.  ah, you say, this isn't relevlant in the bush/kerry case,
because we're looking at the two men in the *same* context, namely a
particular presidential debate (plus the two still to come).

but this is to use the word "context" inappropriately, to pick out only
objectively defined properties of the setting.  the notion of "context"
we want in sociolinguistics is a relationship between settings and
actors in those settings; it's really what actors see themselves as
doing in those settings.  we use the settings themselves as proxies for
the things we're really interested in.  sociolinguists do this sort of
thing all the time, using education level and other measures as proxies
for social class, ancestry as a proxy for race/ethnicity, sex as a
proxy for gender, and so on.  we get away with this, some of the time,
only because the effects in question are sometimes so strong that they
survive this very indirect exploration.  and we do it for practical
reasons: because it can be hard work indeed to get at the
characteristics we're really interested in.

in any case, getting back to the bush/kerry thing, there's a very real
possibility that the context was/is *not* the same for the two men,
that they configure the setting and their role in it in different ways.
  and that this difference in context, rather than some difference
between the two men's presentations of self in general, might be the
source of whatever differences in linguistic usage there turn out to
be.  kerry might, for instance, see his role primarily as a debater,
while bush sees his primarily as a wooer.  (yes, i know, the events are
labeled as "debates", but such labels merely guide, not determine, the
way participants see themselves as fitting in.)  that would lead kerry
towards stressing reason, bush towards stressing emotion and
affiliation (not an implausible suggestion, given, for instance, bush's
references to how hard it is to do the job of president, thus appealing
to his audience to identify with him), and it would move kerry towards
the more formal end of the style world, bush towards the informal and
colloquial.

but enough for a while...

arnold (zwicky at csli.stanford.edu)