I'm surprised no one has mentioned it thus far in this discussion
so allow me to recommend Steve Prince's excellent article,
 
        "The Pornographic Image and the Practice of Film Theory"
 
                (CINEMA JOURNAL, vol. 27, no. 2 -- Winter, 1988)
 
In this piece, Mr. Prince describes research he did for his doctoral
dissertation at the Annenberg School in which a sizeable sample of
porn videos were carefully coded to see if all the offences usually
attributed to porn videos were really present in any significant
quantity.
 
His results may surprise you.  Many, at the time, refused to accept them.
After all, what can that quantitive research stuff, based on extensive
systematic viewing and content analysis,
        -- Derek Bouse