Let me be clear: I'm not categorically opposed to film theory - far from it. My worries about the "return of Theory" at SCMS, such as they were in fact worries, had to do with the conditions of its return. Most immediately, the range of theory seems strangely limited, with the same few theorists getting bandied about, the same concepts repeated.
And, too, there's the difference between film theory (which I take to be any broad reflection on the medium and its representational practices) and Theory (which seeks a philosophy of film experience). Whether or not you subscribe to David Bordwell's polemic against Grand Theory, the warning of C. Wright Mills in his Sociological Imagination (which Bordwell is riffing off on) is useful: theorists have a propensity to fetishize concepts, rather than to use concepts to illuminate an object of study.
Let me take up one concept that seems to be in danger of that trap: affect. I'm open to finding this a valuable concept and heuristic device, but so far am unsure. I don't have a fully developed critique or intervention, but wanted to provide a place holder for thoughts that have been bouncing around in my head. Perhaps at some point I can articulate a more sustained response to affect theory, or perhaps readers invested in the concept can point out what I'm overlooking or what readings I should consult.
1) Affect is pretty much a synonym for emotion, yet presumably as a theoretical concept it would mean something more abstract, in distinction to the everyday understanding of emotional response to cinema. But uses aren't always clear on how affect differs from everyday emotion. Or, if everyday emotion is what's at stake, why use the fancier word?
1b) What emotions are we talking about anyway?
2) Affect seems to dovetail with arguments about spectatorial excess, say Miriam Hansen's notion of vernacular modern or Linda Williams' model of body genre, only the path of argument seems to be reversed. Where Hansen or Williams start with observations about historical spectators (fans who swoon at Valentino, women who cry at Beaches) then proceed to speculation about general spectatorial relations, the new spectatorship theorists often start from general speculation about what spectators necessarily experience when they watch movies.
3) Following from #2: How does one assess what emotions people feel when they watch movies? Is some protocol of evidence necessary to assess how viewers actually emote? (And simply relying on personal response does not sidestep this methodological problem.) It's not that affect theorists never provide evidence for their claims, just that scholars with radically opposed expectations of evidence end up rallying around the same critical concept.
4) Even if the move to affect/emotion - and the move to a subjectivist understanding of film spectatorship - is valuable, there seems to be little methodological reflection on what subjectivism entails. For example, the point of subjectivism in social science at least isn't really that objectivist understandings of the world are wrong, so much as that the construction of the object of study determines the kind of meaning - "close-up" study produces a sociology radically different in kind than sociology "from above." How does affect's subjectivism depart usefully from the subjectivism inherent in British Cultural studies (whose proponents were fans of Garfinkel and Goffman) or reception studies (who have the closest thing to qualitative social science research in humanities film scholarship)? It matters a great deal whether we think affect/viewer emotion/spectatorial excess is a supplement to spectatorship studies and textual models or a replacement for those models. Oddly, a number of affect proponents don't think they need to answer that question.
5) Affect is often understood as a discursive construct, in which the scholars treat references to emotion or emotions as a text to be read symptomatically. Yet it's not clear when this discursive formation reflects a true difference in spectatorial relations and when it is merely discursive.
6) At some point, circa 1990, the discipline decided that the 1970s film theory project - generalizing spectatorship as a point of cinematic address - was too ahistorical in that it did not take into account how actual viewers (or at least their group identity) made sense of films. Affect, as a strain of film theory suited to ths post-cultural studies moment, would seem to open space for multiplicity of responses to film. And it does, since spectatorial relations are no longer understood as mechanical textual effects. But there lies a potential problem. Forget the Early Cinema/Late Cinema thesis: affect theorists can and do find the emotional excess of film experience everywhere, in every genre, in every historical period. If the kind of emotional responses to contemporary action films end up being not all that different from the emotional responses to 1930s Chinese cinema, then we've snuck ahistoricism in through the back door. Conversely, if we're content with a methodology that allows for armchair (i.e. nonempricial) speculation about spectatorial relations, why must the 1970s project remain such a bad object?
I realize these thoughts are scattered and come perilously close to the straw man arguments which bash trends without specific examples. So any examples or counterexamples are most welcome.