An ADS evaluation of dialects in movies?

[Nancy Carol Elliott]

In response to Beverly Flanigan's and Larry Rosenwald's inquiries--

I primarily looked at the rhoticity of North AMerican actors representing
Americans, but a few UK actors were examined for comparison. The locale of
the film plot and prtrayed regional origin of the character didn't end up
being significant, but social class of the role did. The main factors in
variation were time and gender. Women were more r-less than men until the
60's, and in the 70's their average drops below that of the male actors
who are from r-less backgrounds (who preserve their native r-lessness
while everyone else becomes rhotic on film). "Bad girls" stay non-rhotic.
The actor's native region made some difference, but the pattern of change
was still the same - steady decrease (except that women's r-lessness
dropped sharply between the 30's and 40's to match the men's rate and then
didn't decrease at all between the 40's and 50's... 40's wartime equality
followed by a 50's return to elegance?) The same actors (I mean both
genders) in different films decreased their rate as the decades
progressed. They were pretty consistent in films of the same time period,
except when there was a big difference in the status of their role.

Within a film, I found lots of acfommodation by men to the rhoticity of
costars and lots of style shifting in the 60's and 70's (to mark
relationships and emotions) by male actors whose native dialects are
variably rhotic. (The pattern was very similar to the Elizabethan theatre
use of you and thou, but you'll have to read my dissertation to find out
how...)

During the 30's to 50's, men in the films very often became a lot more
r-less in the presence of ladies and a lot less r-less to the fellas. Also
in that time, there's a predominant pattern where the male lead is less
r-less than his female costar, no matter what her % is. (Astaire & Rogers
are a notable exception: his rate is always far above hers, but they
carefully choreograph their r-less rates to her status in their different
films, and he always stays the same percent above her rate.)

As for the Southwestern semi-r-lessness Beverly mentioned, John Wayne's
average rate of 4% r-lessness (that's only five r-less tokens out of 120
possible) in his 30's and 40's films is very, very low. Compared to other
30's and 40's male actors, he's unusual - way at the bottom of the range
(and Astaire is way off the scale in the other direction). Early Westerns
seem to be populated by rhotic characters. Even the women are way under
their gender-decade average. Westerns don't get stereotyped 'Slim Pickens'
speech until later on.

I think I just wrote my abstract. I defend in January (Indiana
University).

Nancy Elliott
Oregon Shakespeare Festival
elliottn at indiana.edu

On Fri, 19 Nov 1999, Beverly Flanigan wrote:

> I gather you looked not only at "outsiders" playing Americans (and vice
> versa?) but also at Americans representing Americans.  Apparently you
> correlated degree of r-lessness with both the region/locale of the film
> plot and the  regional origin of the character in question?  Did you also
> consider social class, education, and age?  (You considered gender, as you
> note.)  The steady decline from 1930 to 1970 would go along with Labov's
> findings on post-WWII New York (for some classes); but what about Boston or
> Deep South settings?  Ray Milland could play a prototypical middle-class
> New Yorker of our parents' or grandparents' generation; Kevin Spacey of
> "Midnight in..." represents a younger but upper class Southerner
> (post-'70s, of course).  And John Wayne's 4% r-lessness would be reasonable
> in light of Hartman's (19??) suggestion of semi-r-lessness in the Southwest
> (Wayne's cowboy country).  Only 3 American "authentics" (my term) seems low
> to me.  But I'm assuming you looked at all these sociolinguistic
> factors!  Where did you do the dissertation?
>
> At 09:41 AM 11/19/99 -0500, you wrote:
> >I'm just finishing my dissertation on the subject of rhoticity in American
> >film speech from the 30's to the 70's. I collected data on about 50
> >subjects per decade (4-year period in a decade's midpoint) and got a
> >decade average from each subject's percentage of r-lessness. The decade
> >average decreases steadily from the 30's to the 70's (59 -> 43 -> 33 -> 22
> >-> 7), with interesting differences between female and male subjects.
> >
> >(Correction: Ray Milland's r-less rate is 92, not 98. And John Wayne's,
> >incidentally, is 4.)
> >
> >Nancy Elliott
> >
> >On Thu, 18 Nov 1999, Beverly Flanigan wrote:
> >
> > > Very interesting!  Where did you get your percentages on r-lessness, and
> > > have you published anything on your 40-year analysis?
> > >
> > > At 04:51 PM 11/18/99 -0500, you wrote:
> > > >Anthony Hopkins only gets half a star for his accent as Nixon.
> > > >
> > > >Of the American films I've studied (1930's to 70's) that have overtly
> > > >specified regional origins for a character, I found only three (US) actors
> > > >that made any kind of attempt to sound like that was where they were from.
> > > >
> > > >Of the Brits in American roles, Angela Lansbury is extremely successful --
> > > >of course she came to NY to study acting at the age of 15. Stephen Boyd
> > > >(Fantastic Voyage) gets 4 stars; I only heard a couple of Belfast vowels
> > > >from him. Ray Milland (The Lost Weekend ) didn't sound Welsh in the
> > > >1940's; the only remarkable thing I found about his speech was that he was
> > > >98% r-less, compared to an average of 42% for male American actors of that
> > > >decade.
> > > >
> > > >And then there's Cary Grant...
> > > >
> > > >Nancy Elliott
> > >
>