Folks, just so you know the sacrifice I'm making, I'm in NYC and supposed to be reading Schelling before heading tonight to the Spotted Pig to meet Yana and her friends, but as it happens the blogging conversation demands a response sooner than later (perhaps this means I should wake before noon ...).
Let me head off a few accusations.
Leigh, first, diegesis vs. mimesis equals morality vs. amorality. No. Never meant to imply as such. Completely unrelated beasts, as far as i am concerned. But I may admit that perhaps I am using the word diegesis in a technically incorrect manner (for a more technical definition, may I suggest the following venerable source of information). Primarily what I refer by this is the following: in film narratives one typically finds all of the components of the narrative to be self-referential and self-enclosed. So there are rarely moments in the film that are not somehow pointed to the central narrative that is taking place. The instance par excellence of this is the introduction of an object that will later have significance in the narrative. In an action film, someone will talk about the function of some tool or instrument which later will be the key for the indivdiual overcoming some threat or something. Thus, what I take to be non-diegetic in Goodfellas is the scene where they have a late dinner with Pesci's character's mother and talk about the painting. The painting is pure diegetic excess: it does not really relate to the action that is taking place or contribute to our knowledge of the characters. It is really just a funny incident, an aside, or, for your Derrideans, a detour.
Clearly there are a lot of reasons why we might be skeptical about my definition. For example, doesn't the hermeneutic method of the viewer require that she integrate all events within the film to the central Sache at hand? Wouldn't this then abrogate any such judgment? Moreover, as you know some of my theoretical prejudices, wouldn't this be a violation of the non-intentionalist readings of films? Sure.
However, concepts are ready-mades, useful in some situations and not in others. In general, it seems to me that filmmaking is governed the rule of efficiency and against non-diegetic excess. Perhaps there are purely pragmatic, financial reasons for this. But those perhaps purely pragmatic reasons also correspond to the demands of a lot of viewers, namely, that narratives make sense without too much monkeying around.
Obviously a film that is purely non-diegetic would be very possibily not a film ... but video art? Yet there is soemthing to my mind disturbing about the demands of diegesis such that a film bears no consciousness of its own production and its own possibilities. I wish I could explain why I think that is, but it seems it would get me too involved in questions that I cannot yet answer. Nonetheless, it seems that film should open up questioning. Perhaps I can make this normative demand of all film. That film that does not provoke speculation is meaningless. And I think one of the gestures par excellence for doing that is by the careful use of non-diegesis.
Does non-diegesis then mean "mimetic responsibility," as Leigh claims? I think that the beyond to which I refer is not the world outside of the film, but merely an interaction with the narrative outside the terms that narrative establishes for itself. Again, I can think of numerous objections to this claim. For example, that a film refers to itself, for example in Funny Games, with the use of the remote, nonetheless makes self-reflection a part of its own narrative. Thus, any gesture to provoke reflection would become diegetic.
(I am really wondering if the opposition posed between diegesis and mimesis in the entry I linked above can really stand in any creative work.)
But I want to stick to a simple sense of diegesis: that of a plot with a beginning, middle and end; a introduction of characters, production of a crisis, and a resolution of the crisis; etc. In other words, the most basic rendering of a narrative form.
This being said, I want to recast my comments about Scorcese versus Anderson. I don't fault Scorcese for his own particular sense of diegetic responsibility. Obviously Scorcese is a master at the creation of these particular types of narratives. Namely gangster narratives. He has his finger on the pulse of a certain force within American culture. But I think in the end we value directors more for their contribution to the form of film. This is why we think Citizen Kane is more important than Goodfellas. Well, at least the people who do film theory/history. And this is the reason why P.T. Anderson may be a more important director than Scorcese. Frankly, I was underwhelmed when I saw There Will be Blood. But I am convinced, nonetheless, that there is something very important about this film.
One of those things, I think, is the denial of the moral obligations that Leigh imposes upon this film. I hope that I'm not reducing your claim otherwise, except to what you pose it as, in the following: "my criticism of Anderson’s There Will Be Blood really just amounts to claiming that Anderson did not tell the story he could have or should have in that film." Not that you are claiming that "any film that doesn’t include a coherent, didactic and thoroughly sensible narrative structure is a bad film." (Although, I must say, I think your use of the word "didactic" along with these descriptors of cogency seems peculiar, is not self-deconstructive.)
I mean, I think that There Will Be Blood denies, purposely (although this qualifier is not important), any moral imperative. To me, the whole question of the film revolves around the final part. I am not sure why in the final segment we encounter Plainview and H.W. years later, when the latter has grown up and has intentions to set out on his own, and they break definitely. Plainview is clearly mad. Then Eli shows up and we have the great milkshake-straw analogy. Finally, Plainview murders Eli. These two events are in some way the resolution of the two central events in the film: (1) when H.W. is struck deaf and the tensions this produces; and (2) when a piece of land necessary for the pipeline requires Plainview to go and repent and prostrate himself before Eli's church. The encounter with grown H.W. resolves this earlier tension by merely definitively ending their relationship, revealing the illusion of Plainview's paternity. Here we have Plainview reaching the absolute lowest point of his moral character. The last possibility for redemption he repudiates, ostensibly reflecting his vacuous, yet total, capitalistic ethos. In other words, from a moral point of view, Plainview is condemned, reprehensible, monstrous--yet a monster produced by an industry and a specific mode of production (this is, I presume, consistent with Lewis' intention).
But when Eli shows up the moral meaning changes. For Eli has been represented as a small time charlatan, lording his meaningless morality over Plainview and having the satisfaction of Plainview prostrating himself before him. If Eli is the representation of morality apart from the machinations of early 20th-century oil capitalism, there is a tacit smirk in this representation. Regardless of the sanctity of Eli's morality, there is a certain truth to the way that Plainview has simply manipulated the community to which his riches perhaps rightfully belong.
Yet Plainview gets his own revenge for being belittled by Eli's morality. He gets the satisfaction of showing Eli that the latter has no legal recourse to the oil "owned" by the church. As well, he has the satisfaction of revealing the self-serving motives of Eli himself and the superficiality of the Christian morality Eli represents. But that is not enough. Then, finally, Plainview crudely murders Eli, essentially claiming that nothing less than death is the penalty for the injury Eli has done to him. Thus, in short, Plainview has destroyed the only representation of morality, showing his insuperable, albeit mad, power.
Hmm. I know I'm lacking a conclusion. But I am going to stop there for now.