Ha ha, after reading about the Errol Morris non-controversy ginned up by New York Times, was checking out his Times-hosted blog, which is intermittent but extremely wordy. This installment includes a hilarious exchange between the filmmaker and "Dan Levin, a professor of psychology at Vanderbilt University" who has been studying continuity errors in movies. As a test case for such errors Levin points to Luis Bunuel's employment of two actresses to play the same woman in That Obscure Object of Desire. Some people evidently don't notice the switch. Although Levin has shown the film to experimental test subjects, he admits to Morris that he fast-forwarded through it himself "because he is generally not a Bunuel fan" and because "the male-female thing is so aggressive and negative." That's great--fast forward scholarship. Elsewhere in the post Morris tells us he prefers one of the film's actresses to the other, so the substitutions "drove him crazy" last time he watched it, but he coyly doesn't tell us which actress he prefers. (Carol Bouquet is the more magnetic and "obscure" of the two--the other, Ángela Molina, looks like a last minute substitute, which she was. What's the big secret?)