I am currently engaged in defining what I mean by self-reflexivity/metacomics in my dissertation, specifically being able to describe not just what Moor, Gaiman, and Ellis do/did, but what they do/did differently. I have a tentative thought.
Mitchell's idea of the metapicture might be helpful, here. To him, the metapicture is a drawing that that can't be either/or, nor can it be both. The Devil's Fork, for example, looks like it's both two- and three-pronged all at once, but it can't be. It uses tricks of two-dimensional art to create something that appears three-dimensional (akin to suspension of disbelief, perhaps?), but using those two-dimensional tricks creates an object that cannot exist in three-dimensional space. The Devil's Fork, specifically, is neither two-pronged, nor three-pronged, nor both. It is, literally, a mere trick of the eye.
M/G/E's narratives (much more so the first two, unfortunately) cannot function as either narratives of continuity or commentaries on metaphor and perception, nor can they be both simultaneously. They clearly do both things, tell stories and comment on perception, but if, for example, we take the constant reminders that everything's metaphor seriously, which appear in Sandman and Promethea every few pages, then we cannot feel empathy for the trials of the characters; they are merely "ideas cloaked in the semblance of flesh," as it were. Dream's death shouldn't affect us emotionally, especially since he doesn't actually 'die,' but we are affected by it because we've come to have an emotional relationship with the character, however that character has been instrumental in demonstrating to us that he's merely a story, merely a fiction, not 'real.'
I'm trying to illustrate a narrative paradox, here, but I feel like I'm not making myself terribly clear. To boil it down to one statement: these are narratives all about how narratives are contrived, but they nevertheless drag us into their internal continuity. That's logically untenable. The basic elements of the plots simply can't work rationally. However, they do work narratively because that irrationality is, in itself, entertaining, much in the same way that a joke's punchline entertains by virtue of pointing to something that's logically impossible.
In the same way that metapictures demonstrate to us how pictures work (i.e., two-dimensional figures can create images that are impossible in a three-dimensional world), M/G/E's stories demonstrate how narrative itself functions differently than reality (how, exactly, is something I have to investigate further, based on close readings and formal analysis). Thus, we arrive back at the premise that Dr. Barbour provided for me three years ago: these are stories about stories, about how narrative works, what it can do, and how it can utterly demolish our sense of reality.
So that's a strong candidate for the kind of the self-reflexivity that the Brit Pack does. I still need to, as Dr. Bucknell says, connect that back to the basic argument that (a) comics deserve scholarly attention, and (b) these particular comics deserve attention within the context of comics studies. These two problems might be solved by the next set of questions to answer...
The strongest construction that I can come up with to demonstrate how these writers 'revise' Anglo-American comics is not that they were the first to do these things (the history of comics is too big to say that definitively), but that they were the first to do them (a) in this historical iteration of comics (the Anglo-American tradition) in such a way that (b) their methods were so popular and/or influential that they became part of the mainstream, ingrained in the artistic practise of the field.
This argument in turn requires that I (a) explain both why and how their metacomics became so influential (which will simply take research and hard thinking), and (b) show several examples of similar self-reflexivity/metacomics across the spectrum of Anglo-American comics (which isn't tough). There remains the argument that these things have been done before, but, again, a collection of counter-examples will suffice to make the case that haven't been done on the scale that they presently are.
To sum up:
1) Metacomics function narratively like unto how Mitchell's metapictures function visually.
2) The 'revision' undertaken by the British Cohort was to employ metacomics in the Anglo-American tradition.
3) The techniques of metacomics were so influential that they became part of the collection of generic expectations within the Anglo-American comics tradition.
Does that make sense? I'm dancin' pretty fast, here.
Dividing into verbal, visual, and combined narrative methods for the sake of analysis is a great idea. I've also bumped into the idea that we can break the comics page down in terms of design and cinema theory: mise-en-scene and composition (colour, light, setting, costume); montage (panel transition, pacing, Eisenstein's dialectic cutting); and collage (essentially the composition of whole pages). I'm currently reading up on the film end of those ideas (mise-en-scene and montage), and will get to art criticism sometime next week.
I can use the above discrete steps (image, text, and Mitchell's imagetext, as well as mise-en-scene/composition, collage, and montage) as the main parts of my close-reading method, but as you say, starting with close reading is not a way to actually construct the dissertation. My intention has always been to engage in the close reading, collect evidence, and then extrapolate a conclusion from the evidence. But, of course, I would then reframe my conclusion as a thesis in my introduction, as Dr. Barbour says. My methodology will also be in that introduction, which will probably be a cut&paste from the proposal, with whatever changes are mechanically necessary at that point.
Mr. Smylie shows pretty clearly that we can see comics as an interplay between, in cinematic terms, montage (cutting, pacing, panel work) and mis-en-scene (colour, setting, costume, light). But by the same token, that panel work, that montage, can also be looked as collage (layering images, grouping them on a given surface to create collective visual impact), so we're back to design concepts, again, despite a very cinematic 'feel' to comics.
The medium itself seems to be an endless string of formal straddlings, image/text, composition/juxtaposition, montage/collage. It's built on the tension of supposedly opposite formal effects. However, upon further investigation, we often find that those ostensible opposites are merely extreme positions on the same spectrum. Text is by no means absent of imagistic impact; playing with fonts makes that impossible to deny. Text really is just another kind of image, and it's only virtue of using a phonetic alphabet (as opposed to pictographics) that we can be made oblivious to what is, in all truth, quite obvious: letters are drawn.
And that same operation, seeming opposites turning out to be far more closely related than we tend to assume, runs through comics. The composition of individual panels (mis-en-scene) and the pacing of panels (montage) aren't all that different. We can, for example, talk about the mis-en-scene of a whole page; paintings often involves attempts to tell stories, relate plots, through a frame jammed with signifying images. Just add lines in between them, and we have comics. The difference is slight, at best.
Once we add lines, though, and start to layer the individual panels, we're talking about montage, something that is, to my knowledge, not often associated with sequence and plot, but is very much a part of the visual pacing of comics, the layered panels and negative space of Michael Avon Oeming's Powers uses what are a kind of collage effect all the time. And once you call that kind of panel work collage, which itself we're linking to montage, then we're back to the mis-en-scene of the whole page, especially if we look back to Oeming's colour effects, which can be stunning.
Comics have the appearance of different media in tenstion (Michell's image/text), but they can have the effect of demonstrating that there isn't actually a concrete difference between those media (Mitchell's imagetext), and instead that they actually bleed into each other in exactly the same way that genres bleed.
I haven't read any of this series and thus far all I know is that it's "a pagan Joan of Arc story," and that it's structurally inspired by Conan as "a fantasy biography." I'm also told that the sex scenes border on porn and action is bloody as all hell, but that's beside the point. Here's an excerpt from an interview with Smylie about his book...
ST: Artesia has an intensely cinematic feel, particularly in the battle scenes. How do you incorporate cinematography and film-making sensibilities into the world of sequential art? Do you have any experience with film?
MS: I sometimes think of what I do as closer to storyboarding than traditional comic book art, but that's probably just because I'm not as inventive with my panel use as other artists. I tend to imagine a physical space and then position the characters in it, and then I'm basically positioning a camera in that space, picking an angle and framing a shot. Tony Caputo and I had some discussions about applying film-making terms to sequential art that wound up in his Visual Storytelling book; we both agreed that you can take the two traditional cinematic styles, montage (the cuts between camera angles and scenes) and mise-en-scene (the arrangement of objects and light within the frame itself) and apply them loosely to comic book art. I think some comic book artists take after montage film-makers, and are most interested in panel use, which determines pacing in comic book narrative; they play with panel size and placement, and tend to be most interested in creating a sense of action. I tend to think most American comic book artists, influenced by Kirby, and Japanese manga artists fall into this category. Other comic book artists are more mise-en-scene artists, interested more in the composition of what's within the panels, and I think European artists are the best example of that; they tend to use fewer panels, and each panel is approached almost as an individual piece of art (and I think that's the style I follow). Of course, I guess the best artists would be those that combine the two styles, and use both montage and mise-en-scene techniques equally well; Mike Mignola comes to mind as someone who both has a great sense of page composition (panel-to-panel narrative) and internal panel composition (with a great formal sense of structure).
Smylie's articulating, from an artist's perspective, the two dominant art forms that we find in comics, the single image (painting, drawing, photography) and cinema. Arguably, comics, in one form or another, have been around for about as long as the single image. If we take McCloud's very stripped-down definition of the medium, the second that anyone anywhere put a second image up next to a first image and they were viewed in sequence, we had comics (though I still argue that that's merely an example of sequential visual art, and that comics as such didn't come along until much later). Comics have definitely been around longer than cinema, but cinema is a much more influential art form, so despite the chronology, it's perfectly reasonable to talk about the form of comics as having been heavily influenced by film.
So we have two basic formal features, here: composition and sequence, what Smylie calls mis-en-scene and montage. Montage is just the French word for 'cutting,' but intellectual montage, as theorised and practised by post-Revolutionary Russian film-makers, involves a particular kind of cutting, the dialectical 'collision' of images. The juxtaposition of images on the screen creates, through that juxtaposition, a third concept; thesis and anti-thesis combine to create synthesis. Arguably, this is the central conceptual mechanism of sequential art, at least in the Eisner/McCloud model. Even the simplest action-to-action sequence requires that the viewer fill in the movement between the first panel and the second, McCloud's "blood in the gutters." That filling in of movement is a synthesis of the first panel and the second, and it doesn't, as both McCloud and Eisenstein (the most famous proponent of intellectual montage), actually exist either on the page or the screen. The 'intellectual' part of the montage exists only in the viewer's mind. Composition/mis-en-scene is, as Smylie implies, almost the complete opposite. It's the arrangement of visual elements on the page or the screen. Altough it, too, can evoke ideas outside of itself, it's power is lies in what's there, for the most part, not in what's implied by comparison.
The difficult part is that mis-en-scene is a lot easier to separate from montage, in practise, than composition is from sequence. Each comics page has its own total mis-en-scene, which is made up of individual images. Conversely, a 'single' image can set up internal sequence and juxtaposition that creates the same dialectical process in the viewer's mind. I'm sure an expert in film theory could point out more complex, meaning murky, implications of constructing mis-en-scene and montage as diametrical opposites, but we're actually talking about comics, here.
This leads to the not-terribly-obscure idea that to study the formal qualities of comics, I really need to study the formal qualities of both the single image and the cinema.