So I wrote Kristin Thompson. Here are two relevant pieces of her email reply:
I’d say point them to the paragraph on p. 92, “Flashbacks offer a fascinating instance ...” Basically it’s a convention of classical cinema that flashbacks from characters’ point-of-view actually quickly slide into giving spectators information that the narrating character couldn’t know. I guess I’d say that in a film like Sunset Boulevard, we have the illusion of subjectivity, but in order to tell the story fully, the filmmakers slip into a largely objective approach for most of the flashback.
This way of thinking about the relationship between subjective and objective narration holds true for Double Indemnity and Out of the Past as well. The idea is that the film mobilizes the mode of subjectivity in order to access the past, but once there reverts to objective narration of that past for the viewer.
In a different email, she added this:
Thinking a bit further about flashbacks and subjective presentation of narrative information. Another convention is that Hollywood films usually provide some pretext for moving into a flashback. Sometimes there’s just the wavy lines or the dissolve or whatever, and we’re in the past. More often, there’s someone starting to tell other characters a story of the past, a shot of a character remembering, a diary page, or some such device that motivates the flashback. So subjectivity is an excuse to move into the past rather than strictly a way to move into the mind of a character and stay there. And even if a character’s telling or memory is the motivation for the flashback, we are to assume that everything shown in a flashback is objectively true unless there’s some further indication that it isn’t. When a flashback doesn’t show the truth, as in Stage Fright, people tend to think of it as cheating. That’s not logical, since we know this is a character’s story or memory and could thus be inaccurate, and yet we’re so used to essentially objective flashbacks that we see any deviation from that (unless signaled to the audience) as sneaky on the part of the filmmaker.
It is useful to remind ourselves that much of what we see in film is rooted in conventions developed over time (like the notion that what we see in a flashback should be the objective truth), but really aren't set in stone in any meaningful way other than that they are usually followed.
She also added this
another thing I would stress is that the concepts and terms we introduce in the book are analytical tools for us, not rules that filmmakers follow. Conventions have been developed for storytelling that allow them to mix the ways they give us narrative information, but obviously they do so in ways that usually don’t allow us to notice inconsistencies—unless we’re writing term papers and the like.
which is useful for reminding you of something I've mentioned in class at least once if not more, that these concepts and categories are tools for you to use in breaking open a complex film in order to understand it. If a particular textual move -- like having a character initiate a flashback subjectively that is then presented in objective narration -- seems awkward to you, don't try to force it to fit into one or the other set of categories. Instead, use your own analysis of the film and the categories you do have to explain how you see the film working. This is really the best we can do, but with the best film analysis, it is enough to open up entire new perspectives on those films.
What are you thoughts on this way of studying complex art forms like cinema? Is there something lost in approaching films this way? Would prefer our discipline to be more rigorous? Do you see any value in approching films this way?
