Monday, November 28, 2011

Another statistics/auteur theory post -- causality

[the second in a very infrequent series]

From the LA Times:
Cinema trends ebb and flow, but one facet of Hollywood moviemaking proving remarkably consistent is gender inequality, according to a study being released Monday by USC's Annenberg School for Communication & Journalism.

In a survey of the top 100-grossing movies of 2009 — including "Transformers: Revenge of the Fallen," "Harry Potter and the Half-Blood Prince" and "The Twilight Saga: New Moon" — researchers found that 32.8% of the 4,342 speaking characters were female and 67.2% were male, a percentage identical to that of the top-grossing movies of 2008.

"We see remarkably stable trends," said USC Annenberg associate professor Stacy L. Smith. "This reveals an industry formula for gender that may be outside of people's conscious awareness."

...

Researchers found that the sex of the storytellers had a significant effect on what appeared on-screen. In movies directed by women, 47.7% of the characters were female; in movies directed by men, fewer than a third of the characters were female. When one or more of the screenwriters was female, 40% of characters were female; when all the screenwriters were male, 29.8% of the characters were female.


The article doesn't quite come out and say it explicitly but I think it's fair to read this as claiming that there's a causal relationship between director's gender and the number of female characters, and that's really not feasible.

Gender of characters with speaking roles (which seem to be the ones we're talking about) are specified in the screenplay and, despite a widely held opinion to the contrary, directors don't write screenplays. This leaves us with two possibilities: writer/directors are confounding the data or the causal relationship runs the other way -- a screenplay with more female characters is more likely to be directed by a woman (perhaps because female directors are seen as less capable of handling male-heavy action films).

This question of causal direction has interesting implications for other auteurist analyses. How many of the common themes we see in a director's work are the result of the artist's personal vision and how many are the result of a certain director being assigned a certain type of picture (director as subgenre)?

Take Dial M for Murder. Thematically, it certainly feel like Hitchcock, but it was a successful stage play before it was a movie and pretty much all of the themes were in place before anyone even considered a motion picture version. In this one case isn't it reasonable to suggest that the themes caused the studio executives to pick Hitchcock rather than Hitchcock being responsible for these themes?

No comments:

Post a Comment