This is no sh*t.

Arnold M. Zwicky zwicky at CSLI.STANFORD.EDU
Tue Aug 3 18:20:33 UTC 2004

On Aug 2, 2004, at 7:07 AM, Jonathan Lighter wrote:

> Arnold's long and considered response to my expression of
> impressionistic opinion valuably helps clarify the entire issue about
> langauge and gender.
> I must return, however, to the basic question, which is intended to be
> not offensive but provocative:  Where is the DATA to show that women
> use as much or more slang than men - or vice versa?  Arnold's argument
> implies that the "correct" a priori position is that the sexes are
> quite equal in their use of slang.  But there is no correct a priori
> position in a question of fact. I simply don't understand Arnold's
> interpretation that JUST ASKING the question in the way I did is to
> affirm the Flexner position.

Well, you said "This is also my impression."  Sounds like an
affirmation to me.  But to the core of the matter:

>   So how about this?  Where is the DATA to show that men use more or
> less slang than women?

My posting was also intended to be provocative, by suggesting that
women might use as much slang as men or more; after all, there's as
much evidence for this as there is for Flexner's claim.  (And being
"correct" has nothing to do with it, assuming that this is a swipe at
me, suggesting that I am merely unthinkingly parroting some party
line.)  But in thinking about the question, my opinion has changed.
When I started writing my posting, my intention was to end up with a
simple Not Proven, plus some serious misgivings about whether the claim
could in fact be examined empirically, plus a conviction that the claim
is actually of no interest.  (Suppose we somehow managed to bring data
to bear on the question, and we ended up with a result much like that
for height -- with men, on average, using a significant, but small,
amount more slang than women, on average.  Why would that be an
interesting result?  What would that tell us about the nature of men
and women?  Nothing, it seems to me.)

But by the time I'd finished the posting, I was well along towards the
conviction that the claim was not only uninteresting, but in fact
unanswerable.  And that's where I am now.

First, however, some anecdotal evidence suggesting that the claim is
true, and some suggesting that the claim is false.  On the true side,
there are the reports of several people on this list (and many not)
that, in certain contexts (involving particular subgroups of modern
American speakers of English) familiar to them, men seem, to them, to
use more taboo slang than women (both in types and in tokens).  I have
no doubt that there are such contexts, and I imagine that these have
been studied, possibly with reference to a single taboo word (like
"fuck" and its variants) or small sets of them.

Here I would like some references.  I'm by no means an expert on taboo
words, or on slang in general, and I'm ignorant of the literature.
Still, I'd be astonished if no one had looked at gender differences in
the use of certain taboo vocabulary, in some population with other
social variables (like region, class, age, and so on) controlled, and
with context also controlled.  (It would scarcely do to compare
working-class teenagers hanging out with one another to
upper-middle-class professional women addressing an audience.)  As I
say, I have no doubt that some such studies would show a statistical
advantage for the men; after all, taboo vocabulary is associated with
masculinity for many people.  I also have no doubt that particular
populations and contexts can be chosen so that a study would show no
statistical difference, or even an advantage for the women; after all,
taboo vocabulary is associated with other things than masculinity.
(There might well be no such studies in the literature, though.  We
tend to look in places where we'll find what we think is there.)

On the false side, consider the usage of "social slang" (evaluative
slang denoting 'very good', 'very bad', 'disgusting', 'stupid',
'out-group', and the like) by present-day middle-class American
teenagers.  Recently I've read or heard several commentators (all male,
but not all white) talking about such social slang (unfortunately, I
didn't keep a record of these remarks, not realizing that this would be
a topic I'd have some interest in).  When they wanted information about
what's the in thing to say these days, they consulted...their teenage
daughters.  Who supplied them with elaborate lists of terms, their
gradations of meaning, and sketches of their recent history in their
peer groups.  Cool, baaad, hot, grody, icky, squicky, retard, feeb,
dweeb, nerd, on and on and on.  Try this with a teenage girl of your
choice -- even better, with several of them who are friends.  Then try
it with a teenage boy, or boys, of your choice.  I bet you'll get  a
lot more response from the girls than from the guys.  Then do a study
of the use of social slang under suitably controlled conditions (as
above), and I bet you'll find the girls using significantly more of it
(in types or in tokens) than the guys.  (I'm hoping that someone has
done this, but maybe not.)

What you find depends on where you look.  Language use, here and
everywhere, is exquisitely context-dependent.

Today's reading, from "Relational practice in the workplace" by Janet
Holmes and Meredith Marra, in Language in Society 33.3, which
(conveniently) arrived in my mail yesterday, p. 385:
Is humor gendered discourse -- discourse that contributes to the
construction of a person as male or female?  And if so, with which
gender is humor associated?  On the one hand, there is research
evidence that women at work are regarded as lacking a sense of humor...
  On the other hand, whenever we discuss such research, especially with
working New Zealand women, they are at least skeptical and often
incredulous of such claims.  Our own analyses of humor in the workplace
provide substantial evidence that, in a variety of contexts, women
contribute at least as much humor as men...

Back to Flexner's claim.  It's a claim about all people, generalizing
across all types of slang, times, places, social groups, and
*contexts*.  To examine the claim, you'd have to estimate the frequency
of slang use for randomly chosen males and females, with all
interfering factors removed.  This is a statistical nightmare.  You'd
need to have an inventory of relevant contexts (this is
culture-specific, and constantly in flux), so that you could sample
contexts randomly, and some estimate of their frequency of occurrence,
so that you could correct for effects where very frequent contexts
swamp others, and some estimate of the availability of these contexts
for males and females, so that you could correct for differential
access, and some estimate of the occurrence of extreme types ("men's
men" and "schoolmarmish women", say), so that you can correct for the
bias they might introduce, and some estimate of the frequency of
various types of slang, again so you could correct for *their*
frequency biases.  There are probably more things I haven't thought of.
  I simply can't imagine taking on a project like this.

Nor can I imagine why anyone would want to.  The point of it would be
to estimate some base rate of inclination towards slang use, for people
in general (well, let's say, present-day speakers of English) and for
males and females in particular.  What on earth would these base rates
*mean*, socially or psychologically?  Nothing that I can see.

Now, I can see a real future for group-specific and context-specific
studies (comparable to the Holmes & Marra studies alluded to above).
But the general question, it seems to be, is of no interest, and
probably not answerable anyway.

arnold (zwicky at

More information about the Ads-l mailing list