Clement and Distant Reading

I was fascinated by the Clement article and by her approach to The Making of Americans.  I’m not familiar with Stein’s story or the computer programs Clement used, but it seemed that the text provided the methodology.  Clement, pointing out what seems to be the accepted understanding of Stein’s work, says, “the repetitive form… renders the reader’s usual process of making meaning useless and emphasizes the fact that ‘Sense-making is a fundamentally cultural activity’” (362).  Later, she argues that Stein understood her own text as “foreground[ing] the process of meaning-making rather than meaning itself.”  The article, in many ways, encourages the process of making-meaning as well.  I couldn’t help feeling that Clement’s article was defending “distant reading” through the analysis of Stein’s story.
 

Just as The Making of Americans has two halves (one traditionally linear and one repetitive, “mimetic reminder of the impossibility of exact replication”), reading similarly has two-halves, close and distant (376).  Clement argues that the D2K application discovered patterns from the “chaos of the more frequent repetition,” and shows textual constructions that “may have [been] missed with close reading” (363).  The overall effect of distant reading and close reading Stein’s text “leads the reader to consider that the formation of knowledge is a cycle of the ongoing creation that results from this push and pull” of the “hole of knowledge” and the “whole of knowledge” (376).
      

It seems to me that close reading and distant reading provides a similar method of the formation of knowledge conveyed in Stein’s work.  Moretti says distant reading is a “specific form of knowledge,” suggesting that it provides a unique approach or specific vantage point of text.  I’m interested in the preparatory work done before initiating a computer program.  Clement decided to analyze the frequency of repetition in Stein’s work, and then map the “co-occurrences.”  This decision seemed to be based on the background (the critical complaints) of the text.  I want to know what happens to analysis when one algorithm is privileged over another, one motif or technique studied instead of another.  Does distant reading focus too narrowly on the pre-set conditions of the computer software?  Or, are these programs more efficiently mimic the way we read?  I’m interested in what would happen if we distant read A Portrait, and whether it would complement or disagree with our readings.

 

Comments

 I am also pretty interested in the substrate levels that work on texts like the algorithms and so on.  This seems kind of ridiculous, but does an algorithm that is applied to a text becomes a text itself?  How about the product of the application of the algorithm?  Is a meta text or critical findings an iteration of the original text?  In other words, are these formalist analyses just different ways of looking at the same object?  This kind of goes back to the ontology of the text.  What is The Making of Americans?  If it is rearranged or data mined does it cease to be itself?  Or, if we saw a graphic representation of the whole text, is that something other than the text itself?