Tag Archives: Franco Moretti

Graphs and Code

11 Feb

The promised follow-up to my previous post on Paratexts will have to happen on a later date.  For now, some thoughts on someone else’s thoughts on quantitative data and the history of literature

One thing that continues to dazzle me about the humanities generally is its seemingly limitless capacity for appropriation.  It would seem no discipline is safe from students and scholars of literature.  Freud, Foucault, Marx, every manner of philosopher, sociologist, linguist, psychologist, anthropologist and cultural critic has had their ideas incorporated into critical inquiries into every type of cultural object by people in English departments in universities around the world.  I must admit, Thomas Kuhn is one of the last philosophers I expected to see name-dropped in an essay about the history of the novel.  What is more impressive is that said name-dropping was not at all gratuitous or misplaced, as near as I can tell.

Franco Moretti is conscious of the gap between text and interpretation to a refreshing degree.  In his chapter “Graphs” from his book Graphs Maps Trees he foregrounds three models of historical inquiry (philosophy of history is a major theme of his book as it is principally concerned with the now-neglected field of systematic literary history) as well as one of the three models of information that make up the title of his book.  The three models of history are Event, cycle, and longue durée.  Event refers to the focus on the specific, individual-centered historical occurrence, while longue durée is principally concerned with, in Moretti’s words, “the very long span of nearly unchanging structures.”  He posits that literary studies have been comfortable with these two historical models generally, and that they figure prominently in literary scholars’ work since the Event model is helpful in isolating the specific circumstances of the production of studied texts and the longue durée is applicable to more ‘theoretical’ inquiries that allow scholars to track long-term and cross-textual and cross-cultural patterns.  The middle-ground Cycle model has been largely neglected.  The Cycle model can probably best be understood- again, quoting Moretti himself- as “temporary structures within the historic flow.”  It is a model which can be used to quantify phenomena that last for a certain period of time but then cease.  The best way to exhibit such phenomena is, of course, via a graph.  This is where Moretti makes a crucial observation- that quantification and interpretation are separate and distinct tasks, and the graph is is only conducive to the former.  I have often been struck by what many consider the central dilemma in the Digital Humanities- The Humanities is principally concerned with interpretation, and it is still not entirely certain how digital technology has altered the process of interpretations, either for literary critics and scholars or people in general.  It is possible that digital technology is merely a tool, one that can be used to gather and quantify information for the student or scholar, but something that does not fundamentally alter his principal task of interpretation.  The precise nature of the epistemological shift brought on by the digital revolution is still being debated, meaning any person working within the field of the digital humanities is essentially required to perform multiple projects simultaneously.  S/he needs to demonstrate the efficacy of digital technology in the study of literature and culture, contribute to the ongoing discussion about the epistemic shift that has occurred as a result of the digital revolution and information technology, and define the parameters of a new literary paradigm vis-à-vis digital technology.  Moretti’s project does not address these problems as such, but can still be counted as an important contribution to the discussion for the simple reason that it addresses the field’s problem with information, the gap between the data and the interpretation of the data.  Principally concerned with the latter, those involved with the humanities occasionally fail to recognize the autonomy of the former.  

In “Graphs” Moretti utilizes his Cycle model for a consideration of the history of the novel, from the 18th century through the 19th.  He discovers that different ‘genres’ (which he persuasively defines as ‘temporary structures’) would appear on the scene, last for a period of 10-30 years, and then be replaced with other genres.  He speculates that the disappearance of one genre and its replacement by another reflects changes in the needs of the culture at a particular historical moment- some genres are more effective at communicating and eliciting certain emotional states than others, and some emotional states are more desired than others at specific points in time.  Anxious times often increase the demand for anxious writings, and so on.  Similarly, Moretti posits that the sudden disappearance of a large number of literary genres is probably the result of the passing of a generation of readers. This is where Moretti’s inventive use of Thomas Kuhn comes in.  These generic shifts- very much contingent on who is producing literature and for whom- result in temporarily stable but contingent literary paradigms that shift when new creative models intrude onto the scene and effectively force out old genres that no longer fulfill the culture’s desires.  Thanks to the tool of the graph, this phenomenon can actually be quantified  to some extent.  And who knows?  Perhaps predictions could be made.  A much larger body of data would need to be amassed for such an endeavor though, and perhaps only an advanced computer program would be up to the task of quantifying all the pertinent data.

Another essay which recently came into my purview was “The Esthetics of Hidden Things” by Scott Dexter, available in the anthology Understanding Digital Humanities.  It is primarily interested with exploring how computer code has altered (or perhaps only revealed) human conceptions of textuality, and by extension, human conceptions of information.  The rhetoric of concealment and revelation figured prominently throughout, and the thrust of the piece seemed to be that the distinction between the hidden and the revealed is very unstable in the digital world.  Because this essay is grounded in a technical literacy I simply do not possess, I do not feel very confident speculating about the merits of Dexter’s argument.  One thing I do find quite remarkable however, is the implication an unstable hidden/revealed distinction may have for the task of interpretation more generally.  It would force any would-be interpreter to abandon the notion of approaching a text in search of ‘hidden’ patterns or motivations.                 

Much else could be said, but time, as they say, is a-wasting.  For now, enjoy some Pere Ubu.   

 

Advertisements