Be forewarned that this is at once the least formal and the most significant part of the theory. In Ulmer's terminology, it is the inventio.
After reading Calvino, I was struck with the idea of a false book. Actually, I was struck with the idea that every book is false. Every author is a liar: the framework in which I desired to work.
That fiction is fiction and characters are inventions is an old, almost trite, idea in literature. It is fairly conventional nowadays for the author to step into his own story, undermining it, explicitly refuting it, and combining in one narrative space the fictive and the real.
A canonical example may be Miguel de Unamuno's Niebla in which the author steps in and murders the protagonist. In Breakfast of Champions, Kurt Vonnegut continually writes, these are characters and I do not pretend that they are anything else. Calvino begins his story with the words, "You are about to begin reading Italo Calvino's new novel. . ." In each of these cases, and in every other example I can think of, however, the nullification of the character is explicit. The author, or his or her fictive personification, speaks directly to the reader.
Although it is an effective method of foregrounding the apocryphy of narrative, the introduction of an author-character into a fiction is somehow contrived. It is the "real world", "truth" invading and condemning (or perhaps justifying) the fictitious world. I believe that a true undermining of a narrative must be implicit; it must come from within the narrative space, not from without it. My suggestion that all characters are unreal is all-inclusive. It is not only true for characters that are denied by the author, but also for ones that he or she sincerely believes in.
Writes Primo Levi in the afterword to The Monkey's Wrench, "With this quotation [of Joseph Conrad speaking of his own character, Captain MacWhirr] I would like to tell the reader that, just as Conrad never saw Captain MacWhirr 'in the flesh,' so I never actually met Libertino Faussone. Like the British captain, Faussone is imaginary but 'perfectly authentic,'. . . ." Authentic, my ass! He is fiction. He's a lie.
Perhaps it is now clear why this theory is not essential (it is, in fact, detractive) to my story and why the story should be read before, and perhaps exclusive of, this "essay."
So how does one question the veracity of one's narrative implicitly, from the inside? By blatant contradictions. By unresolvable characters. By meta-meta narrator disavowing meta-narrator disavowing narrator disavowing protagonist. And what more perfect medium than hypertext, where voices can thrive individually and coexist with equitable veracity.
So whence art history? The Unamuno/Vonnegut/Calvino technique of injecting reality into fiction reminds me strongly of Magritte's well-known painting Ceci n'est pas une pipe, which depicts a pipe with the title written in huge cursive underneath it on the canvas. The effect of this painting is the same for art as the author-character invasion is for fiction. It problematizes the veracity of art. But, also like the author-character, it is contrived; it is an invasion. The text invades the picture space to undermine it; it is the imposition of an exterior medium to voice the problem. Still, a painting of a pipe without such an external nullification would most likely still be seen as a pipe.
There is an art which contracdicts itself implicitly, in the manner that I wish to do in fiction. I refer to the apex of Analytical Cubism, the series of paintings following Le Portugais (1911-12) by Braque and Ma Jolie (1911-12) by Picasso. These paintings utilized the technique of painting flat lettering on the picture plane, on top of the fragmented Cubist space. Note that the letters did not spell words of any significance (to this theory) but their inclusion itself achieved the effect.
Because of the lettering, the paintings contined a synthesis of two distinct dimensions, the third and the second, in addition to Cubist fragmentation. This is not the same as adding a dimension to a circle and getting a sphere or a cylinder, but is more related to viewing the circle and the sphere simultaneously. The two dimensional factor was the lettering itself; flat and without any semblence of linear perspective, it brought the picture plane, the fundamental characteristic of the art of painting into the broken three-dimensional Cubist space, where it hung, a constant reminder that painting is but painting.
A similar effect was achieved by Leger in, for example, the big painting hanging in the Guggenheim (I think it's called Dejeuner ). His carefully modelled personages and flat Cubist planes are washed over with streaks of primary colors that are not involved in the scene itself. These streaks of color foreground the picture plane in an otherwise self-contained narrative painting. Again, the reason I stress Analytical Cubism over Leger is that the fundamental break-up of space so important in hypertext is all but lacking in the latter's work.

Ma Jolie