Sunday, March 3, 2013

More on Moretti: Distance and Mapping

I found this week’s reading on maps fascinating yet perplexing.  The excited English nerd in me who loves to make character maps while reading complex novels was completely on board: “Maps are a great idea! They can add something to our ‘knowledge of literature,’ as Moretti says!” (35). As I reflected more, though, I became perplexed and began asking, what happens to the space that is not mappable? He elusively mentions how Christmas stories cannot be mapped (?) without offering an explanation as to why.  I began to think of other instances in which space is not mappable. Space is never just a location, space is forever bound with time and consciousness (I’m sure there’s a relevant Bergson quote to throw in here…).  So how does one begin to map out a complex novel like, say, Faulkner’s Absalom, Absalom!?  

When one attempts to map out a storyworld, a world that encompasses so much more than just geographical location, it can never be as simple as placing points on a map.   Moretti makes evident the fact that there are “maps and maps” (meaning maps of real geographical locations and maps of fictional worlds) (63).  Though this is true, he does not seem to take into account the complex spaces of those storyworlds, ones that exist at the juncture of the narrative (or, in Faulkner’s case, the multiple narratives) and the experience of the reader.  There are constructed and forever shifting spaces, unable to be tied down and in a constant process of re-definition through the act of reading.  In my edition of Absalom, Absalom!, a map is included in the introductory material.  This map does in fact recreate the geographical layout of the storyworld, complete with points indicating the location of key plot developments.  This map seems almost facetious and mocking though, as if it mocks our belief that the world of Absalom, Absalom! (a storyworld which extends beyond A,A! to encompass The Sound and the Fury) could be represented on a map.  There are quite a few symbolic spaces that cannot be represented on that map, though, spaces such as Sutpen’s Hundred, a space to which characters return time and again and a space that is always changing, seeking definition.  I think it would be fascinating to attempt to map out the symbolic layers of this space and the storyworld in general—plotting all of the different variations on the spaces, overlaying them on top of one another (I’m sure there has GOT to be a more technical and fancy way to do this than the clear overhead projection sheets I’m envisioning).  

To me though, this task is very tightly wed to close reading. It’s simply, as Moretti defines his own maps, a complex “stylization of space” (42).  This brings me back to the point I made in my previous blog post in which I pointed out that, in his early work, Moretti is hardly conducting any sort of distant reading.  What remains perplexing then, is how his methods are taken up (by other scholars as well as contemporary-Moretti) in current DH distant reading approaches.  What Moretti does in the Maps section of Graphs, Maps, Trees seems worlds apart from what Matthew Wilkens does in “Canons, Close Reading, and the Evolution of Method” in which he plots the geographical locations mentioned in 40ish novels of 1851, 1852, and 1874. Wilken’s task seems so surface level (as it’s intended to be).  Yes, this might lead to a “revised understanding of American regionalism” (253), but I feel like this is doing a vast disservice to the texts.  Perhaps it is elucidating something about the canon (though, we would have to extend Wilkens’ sample size) but I believe it’s getting us so far away from what Moretti was doing when he was looking for interesting narrative patterns.  I guess the real question becomes, when we attempt to expand the canon and pay credence to previously unread texts, is distant reading really the way to do that? I believe if we use maps in the way Wilkens does we relegate those texts to the level of data, one-dimensional plotted points.  If we use maps on those texts in the way Moretti did originally, then we will learn interesting things about texts, but a much smaller number of texts.  How much service does it really do, though, to relegate texts to the sad one-dimensional space that has been derived from number crunching?  The English nerd in me ways to make that Faulkner map (or do that map to another multi-layered text), but at what point does that stop being a valuable task?


  1. I agree that, in some ways, Moretti's practices would seem to involve a great deal of close reading--at least, as he puts them forward in _GMT_. Unsure about what one gets out of the project, too, finds you in a similar boat as William Benzon. What exactly has Moretti traced in his maps? "Moretti's maps tell us something about how the mind finds the world. But just how is it that the world makes its way into the mind there to be transformed into texts? That process remains invisible" (Reading GMT 62). Though he does not want to make a specific argument RE Moretti, Benzon moves in a similar direction as you do in his not knowing where to move. Moretti wants to start with the geographic and move to the geometric, form to force. You want to start without geography entirely--one might say that this is not even mapping! Can Moretti's maps work without the geography at all? Or have we simply arrived at the diagram?

  2. Your point about how Wilkens' work seems 'worlds apart' from Moretti's is well taken --- I was musing on the distinction between the two in my own post, though I feel that it's more a question of tone (i.e. I feel that there is room to do Wilkens' work in Moretti's model, it's just that Wilkens employs an invasive theoretical thrust that insistently hierarchizes the payoffs resulting from different methodologies). I agree - his map does 'disservice' to texts, but only because he makes explicit that closely reading these texts is not as 'worthwhile' as making the map (this is certainly a capitalist argument about members of the field using their time more productively - those kinds of claims are so few and far between in the field of literary criticism [I wonder what would happen if at Colloquium we brought up the need to 'increase productivity']).

  3. Your post makes me think of another problem with the spatialization of text that neither Moretti or Wilkens seem to care much about: what to do we do with time? Isn't dureé the ultimate unmappable thing? I'm thinking, of course, of our readings from C. Hanson's course last semester. Timelines and spatial representation of temporality erase the kernel of time in the first place. And yet, if we are to take seriously the spatialization of a novel as complex, as you suggest, as Absalom, Absalom! it seems that temporality ought to be incorporated somehow. Perhaps it is the temporal that drives Moretti to reject Christmastime as a mappable concept.

    For the hell of it, I wonder what we could learn from the hybridized map/trees of skill and story progression in video games. I'm thinking in particular of Final Fantasy XIII's "crystarium." Each character is reduced, abstracted, to nodes of abilities and their interconnected branches. There is certainly something diagrammatic at work here, and in the context of gaming, the notion passes by without any remark. It seems to me that games (not merely of the computational variety) have long been at work abstracting and diagramming representational fictions in order to motivate a different kind of understanding altogether.

    In case you're interested in seeing the crystarium:

    1. Because you implicitly invoked Bergson, I have to reflexively respond with my customary Deleuze-and-Guattari's-Bergson-is-the-key-to-all answer. You brought this upon yourself. Sort of.

      Deleuze and Guattari in both Kafka and A Thousand Plateaus use mapping and cartography as a key metaphor-thingy (not quite a metaphor, as usual for them). But because of Deleuze's work on Bergson, I think they're acutely aware of the role of both temporality and the sort of subjectivity of perception to which you referred in your response to Esei. Their primary move--which I of course think is the right one, since it's the primary move I try to make (I strongly suspect that everyone's eyes glaze over whenever I say "Deleuze" or "affect," anymore)--anyway, their primary move is to map assemblages: in short, the relationship between characters and spaces (or other characters) over time. They draw out Bergsonian intensions and extensions, look for nodes of convergence between lines of flight or exercises of power, and other sorts of jargony things that basically sum up to trying to map texts more spatially than Moretti while taking into account the immateriality of that space and its dependency on time.

      I realize I've basically said nothing other than "Deleuze does it better," but what I'm getting at is that if we want to map both time and space (an inclination that resonates with me), one way to do so is by mapping the relational experience of bodies, rather than plugging in points or rings adjacent lists. It really comes down to how we think texts work, methinks. If one thinks that texts work structurally (or post-structurally, since nobody's yet mentioned Derrida, and we can't have a blog week without Derrida), one might lean Moretti; if one thinks of texts as reader-driven and affect-driven, I suggest one could lean Deleuze and Guattari.

    2. It seems quite relevant to bring up the issue of computers here (which remain starkly absent from Moretti's work). While a "map" in the geographical sense is necessarily 2-dimensional and incapable of representing time, computer generated visualizations have no such limit. It seems entirely probably that we could rig together a digital map which accounts for 4 dimensions (3 in space and 1 in time), or 7 (3 real space, 3 symbolic space, 1 time), or whatever. Of course it seems logical to assume that more dimensions = more hours of work spent on each text--over time we might see diminishing returns. And then we're back to close reading again!