I have been scouting around in the field of “book history,” and it is rich indeed. And of course the rest of the Internet:
[…] [L]iterary work by its very nature sets in motion many kinds of creative intentionalities. These orbit in the universe of the creative work—but not around some imaginary and absolute center. Rather, they turn through many different kinds of motion, at many structural scales, and in various formal relationships. The universe of poiesis no more has an absolute center than does the stellar universe we have revealed through our astronomy. What it has are many relative centers which are brought to our attention by our own acts of observation. The universe of literature is socially generated and does not exist in a steady state. Authors themselves do not have, as authors, singular identities; an author is a plural identity and more resembles what William James liked to call the human world at large, a multiverse.1. Jerome McGann, “The Socialization of Texts,” The Book History Reader, ed. David Finkelstein and Alistair McCleery (London: Routledge, 2002), 42.
Literary texts differ from informational texts by being polyvocal. Whereas “noise” is always a form of corruption for a channel of information, it can be exploited in literary texts for positive results. The thicker the description, so far as an artist is concerned, the better. (Minimalist styles of art thicken their media by processes of subtraction and absence.) A thickened text is a scene where metaphor and metonymy thrive (Coleridge’s “opposite or discordant qualities,” his “sameness with difference”). For [Hershel] Parker, the thickness comes from the artists’ imaginative resources, who can be counted on to put into their texts far more than they are even aware of. Parker’s “intention” includes, crucially, the vast resources of the unconscious and preconscious.
But thickness is also built through the textual presence and activities of many non-authorial agents. These agencies may be the artists’ contemporaries—these are the examples most often adduced—or they may not; furthermore, the agencies may hardly be imagined as “individuals” at all. The texts of Sappho, for example, gain much of their peculiar power from their fragmented condition […] 
[…] But what’s unique in the case of literature is that the text isn’t the finished work. A novel or short story isn’t a solid object. It’s just groundwork, and the reader’s imagination and reference points revive its content, extrapolate from its clues, and finish the work individually. In a way, when you’re writing a novel, for instance, you’re actually writing an unpredictable number of novels at once, and that number depends on how many copies end up being read. No other art form that I can think of involves that level of collaboration between artist and audience and disrupts the passiveness of being a work’s receiver with so much freedom and creativity. The high degree of interaction and codependence that the writer/reader axis makes possible is gorgeous and offers such a great opportunity to experiment with how text reacts when it comes in contact with the imagination. And writing allows you to play with the crapshoot decision of when a work is finished or, rather, when something is ready to be placed outside of your control and then finished by other people in largely unknowable ways. […] 
[…] But texts shape the response of readers, however active they may be. As Walter Ong has observed, the opening pages of The Canterbury Tales and A Farewell to Arms create a frame and cast the reader in a role, which he cannot avoid no matter what he thinks of pilgrimages and civil wars. In fact, typography as well as style and syntax determine the ways in which texts convey meanings. […] The history of reading will have to take account of the ways that texts constrain readers as well as the ways that readers take liberties with texts. The tension between those tendencies has existed wherever men confronted books, and it has produced some extraordinary results, as in Luther’s reading of the Psalms, Rousseau’s reading of Le Misanthrope, and Kierkegaard’s reading of the sacrifice of Isaac. […] 
[…] To this point I have been taking the word “text” to signify the linguistic text, the verbal outcome at every level (from the most elementary forms of single letters and punctuation marks up to the most complex rhetorical structures that comprise the particular linguistic event). And even if we agree, for practical purposes, to restrict the term “text” to this linguistic signification, we cannot fail to see that literary works typically secure their effects by other than purely linguistic means. Every literary work that descends to us operates through the deployment of a double helix of perceptual codes: the linguistic codes, on one hand, and the bibliographical codes on the other.
We recognize the latter simply by looking at a medieval literary manuscript—or at any of William Blake’s equivalent illuminated texts produced in (the teeth of) the age of mechanical reproduction. Or at Emily Dickinson’s manuscript books of poetry, or her letters. In each of these cases the physique of the “document” has been forced to play an aesthetic function, has been made part of the “literary work.” That is to say, in these kinds of literary works the distinction between physical medium and conceptual message breaks down completely. […] 
2. From an interview with author Dennis Cooper that is part of the HTMLGiant “What is Experimental Literature” series of posts. You can read the whole interview (which is great) here.
3. Roger Darnton, “What is the History of Books?” The Book History Reader, ed. David Finkelstein and Alistair McCleery (London: Routledge, 2002), 21.
4. McGann, 43.