The political economy of reading

I've just finished reading William St. Clair's massive new book, The Reading Nation in the Romantic Period, which provides an unbelievable amount of information for anyone interested in either the period or the history of reading in general, along with providing some significant methodological challenges to literary study.
If you aren't familiar with the text, or those like it--which have begun to appear recently, mostly in Renaissance scholarship but also in areas of history of the book, as both of these fields and the discipline more generally supersede historicism and hermeneutics and accommodate sociology--it is a wonderful one that complies a lot of quantitative data on publishing and the consumption of romantic (and pre- and post-romantic) texts. From this data it tries to derive a whole economic system in which they take part (you can get a sense of the assumptions guiding the work from his last chapter, which is in the form of a lecture here). Essentially, St. Clair wants to find out what actually was being read in the period, which, you can imagine, is quite different from what was being written (in the narrow sense of writing) and what was being sold. I can buy a book and never read it; I can just as easily read something very carefully that I see in my passing through a coffee-house.
What keeps St. Clair from being a historicist, however, is what is so crucial about this text: these readings that really did happen only make sense when they allow the construction of a model of the total process of printing, distribution, reception, etc. of which they are a part. What we have here is statistical, sociological, and ultimately economic analysis at its best: it is looking not for the actual transactions so much as the regularities which allow models to be made about probabilities of reading. What happens is that the emphasis falls not on whether something did actually happen--as the historicist loves to point to (I always imagine the historicist as Steve Martin in The Jerk, saying "Stay away from the cans!" when he is being shot at, totally missing the upshot by registering the empirical)--but the network with which we can make sense of the data. This network or system can be refined and recalculated as well as supplemented, allowing us to think off of the model rather than keep trying to pose huge logics or theories about reading each time we do a study. In other words, St. Clair allows me to conclude something different from him based on how I see his model, or how I adjust it: this means that two interpretations are based upon a clear and stable description, and we don't have a new interpretation each time we have new descriptive material given (as in historicist and deconstructive studies). One great outcome of this--like in Franco Moretti's work--is the possibility of serious collective scholarship, which happens already but is, I feel, frowned upon by the particular way our discipline is set up (especially for those writing their dissertations: you have to break out into the discipline as an individual!!).
Most significantly, what happens is we move totally beyond the model of individual agency or influence in the realm of literary production itself: Coleridge depended upon (if one can still call the relationship this--which is precisely what St. Clair upends) hundreds of people in order for his words even to reach their readers. Is it accurate then to say that Coleridge himself influenced his readers to do such and such? Not exactly: the agency is retained at Coleridge's level, but is agglomerated as it also includes all the workers in the field of manufacturing, distribution, and selling, that we have a huge collective producing a text able to be read. The author just becomes one point or node in that collective. What is so significant about this is, however, not that it shows more people were involved than we thought--everyone knows books need to be made, and that there exist communities of readers--but that the issue of influence is spread so far that we have to just junk it. Coleridge didn't influence his readers, because his agency is only so local. A sector of the public is primed to receive what he says, and there is a whole network of technicians willing to take his words up and use them. Pointing to the fact that he wrote something, and particularly pointing to the fact that he wrote it at X or Y time, doesn't mean anything then, because it is not clear whether his words will be significantly powerful enough to achieve anything as significant as "influence" in a system already ready to receive him. One can't say, then, that Coleridge influenced Shelley, say--and one especially can't do this by pointing to the fact that Coleridge wrote before Shelley. It is more likely that Coleridge didn't influence Shelley. Indeed, Shelley did read Coleridge (with Byron, famously, in Italy), but now this term "read" doesn't necessarily mean "influence" or even "determine," in the way that it might have if we didn't know the whole network in which Shelley was placed. At the same time as the study restores a sense of force to the word "influence" and even "agency," it makes us realize these terms are themselves inadequate for describing literature.
Most significant about the book, then, is the way St. Clair outlines the "obsolescent:" literary works that, if we just looked at the chronology, we would not call contemporary with the authors (we think of Godwin's work as contemporary with Coleridge's) and with the public shaped by these authors, but which actually were read. In short, St. Clair outlines how what gets read is usually the books of forty or more years ago, such that one might never have read Godwin or Austen even if one encountered a modern poet like Coleridge--which was rare in and of itself. Reading was stratified into layers of history, and one rarely encountered something that was contemporary. All this makes the assumption that Wordsworth was either directly affecting or (especially) directly responding to his time when he sat down to wrote a poem in 1798, an extremely dubious assumption. If he was responding to texts (and the possibility is opened up that he didn't at all), these were probably those written 40-100 years ago but which he grew up with or read daily. With the less studied receivers of Wordsworth--those who pass by Lyrical Ballads as just another new book in the bookseller's--this is even more the case: what matters for them are the sermons of Hugh Blair, say, or Blackstone's Commentaries, that they read early on and continued to see around them (indeed, Bentham wasn't angry at Blackstone in his Fragment on Government because of Blackstone's arguments: he was angry at how these arguments were in the back of the minds of so many people--he had to read Blackstone as a kid).
But what is so crucial is that even this fact doesn't matter--it isn't that the influence gets displaced over to a level where certain other books, just because they are more popular, still influence people. It is that we have escaped searching for the point of contact between the author and the reader entirely. If it were, then pointing to how a reader was complicit in the system of publishing and receiving something would undermine the system: obviously it would be in his interest to publish one thing instead of something else. But what St. Clair shows is that, so far as the system is concerned, this is perfectly fine:

As simple representations of a highly complex human system, the explanatory power of these models is not undermined by acknowledging that many of the agents were themselves part of the system. Some printers, for example, were part of the constituencies whose ideas they wished to promote. Publishers were themselves readers with their own horizons, needs, and aspirations which they brought to their role as commercial agents.
-The Reading Nation in the Romantic Period, 449

What we need to do is regard things from the point of view of the system, which has--strictly speaking--no interests, no influence, because it is not influential in and of itself. It is only something that makes influence possible, but in such a way that what the system causes is not guaranteed to be influential. In short, it lends credibility to what is not encompassed by the system--and not as an aberration, but as another fact that gives the system itself limits and rigorous boundaries. Other systems can be produced, too, such that one gains a stratified picture reading. One is merely describing the texture of what existed--nothing more. We're not looking, that is, for whether this describes all the readers there are, or the entire system of production: we're simply looking for how much was read and what this reading itself looked like (what were its objects, where did it happen, etc.).
Thus, nothing happens when one points to how a particular agent was complicit in the system. The question for our research shifts: what matters is the shape of the system, the direction that reading practices are heading--in short, how we can make sense of conflicting data. This sense is limited to the coherence of the model, and when the system fails, we can revise it or construct a new one. But what lies outside of the coherence of the model is not a contingency, for everything in the model is contingent.
In short, there is no anxiety of influence: it isn't that it just can't happen (though this itself is really questionable), but that even if it did, it wouldn't matter for what we do and what people did with texts. Complicity is done away with, as well, because it simply isn't significant in the same way: the particular agent that is complicit in the system of production and reception here simply can have conflicting interests that are able, themselves, to be described. What matters is the direction of the interests and whether they were frequent. In the end, everything about our evidence as literary critics changes--and changes, I think, for the better.
Nevertheless, I still have some questions--which I'll end this post with:

1. How portable is this method? Can it work for different economies, especially? I am led to believe that this is the case (one can imagine a similar study in the Medieval period), but that it really isn't much of an objection in the first place. What matters is that it is able to make sense even of the complexities of our system as it is currently--which no other method (even Marxism) is quite able to do with such rigor.

2. Isn't this way of looking at things most helpful in alerting us to unseen or unanalyzed spheres of influence? How can we keep reinscribing this back into the old model of literary history? The method isn't attempting to say something like, literary history works this way only, which would be the only way to avoid this reinscription. But perhaps this is the benefit: other people can argue about what literature and history is. What is important is that we finally have a method that can be amenable to something other than influence.

3. The disposition required to do this kind of work is massively different from that even of literary historians now, not to mention all those studying literature. It isn't a mistake that--like Bernard Williams--St. Clair served in the government and was deft at both using and getting through bureaucracies. Again like Williams, his work takes its form from that experience. But how many people currently in literature departments could do something similar--i.e. would have even remotely similar dispositions? Can you picture your professor also serving on a government committee? I know I can't. Part of this is due to the greater separation of American literature departments from the government than in the UK. But even departments in law and politics have many overlaps with government service. What steps would we need to correct this in the case of literature? Would this not be the real way that literature departments could politicize themselves? Compared with the UK, our departments look quite libertarian. But are they? Is there an important role for this current separation? Are we separate in the first place? And (to conclude this post with a question St. Clair loves) by how much?