Quantifying genre

Moretti's chapter on "Graphs" offered some interesting readings on the "hidden tempo" of literary history (29). The graphs he shows, however, seemed problematic to me for a few reasons, one of which I'll focus on here.

The British novelistic genres graph (19) seemed particularly troublesome. As Moretti says, for his quantifying of the data of genre he "decided to rely entirely on other people's work" (18). Nothing wrong with that, in theory, as that disclaimer goes with the graph and informs one's reading of its meaning. Because of the nebulous nature of the definitions of literary genre, however, this seems to be lacking the most crucial data--the criteria that the individual scholars used in classifying literary history under those genres (i.e. the recurring data, like character types, plot arcs, motifs, etc). What would surely be more useful would be, perhaps, a scatter graph of the appearances of these recurring data that appear in novels from the time period, rather than the genre classifications that were placed onto them afterwards. For instance, having one color dot for the recurring instances of the depiction of the French Revolution, rather than grouping them all in dubious interpretive clusters like "Military Novel" or "Historical Novel"; a colored dot for the character of the "fool"; etc. It would then be possible to identify recurring data more easily, thus enabling the reader, if they wished, to THEN place genre definitions, with greater precision than before. The grouping, by the scholars Moretti consulted, of novels under genres is, I would say, something similar to what Moretti is doing with his graph: trying to interpret literary history in a system. In other words, what a graph of genre seems to be doing is offering interpretations of an interpretation of interpretations.

Comments

 While I think that Moretti's graph on page 19 has some false as far as generalizations go, I do agree that your suggestion could be more valuable. I must note that such a graph would be quite complicated. Perhaps, we will learn more about graphs today, and that will enable us to understand how to crate a graph robust enough to handle that much information while remaining accessible. 

 Just as a general comment, I found this chapter on graphs interesting because in 18th Cent. English Fiction with Dr. Stevens we are discussing the "rise" of the novel but nowhere do we have access to data like this where we can see the geographical and time period information represented.  I find Moretti's ideas useful in a very immediate sense.