The art of graphs

Franco Moretti’s Graphs, Maps, and Trees is a book that exemplifies the vast potential of charting texts in the realm of literary criticism.  It seems to me that in the digital humanities this has become an especially interesting area because with the aid of new technologies charting extremely large data sets has become possible.  One interesting comment that Moretti makes involves the importance of group analysis as opposed to individual analysis (4).  He relates this to the reader in terms of historical studies, emphasizing the potential of literary history as a means to heighten critical analyses through the simplified models of graphs, maps, and trees.  This is interesting, and while I do not disagree with Moretti here, it is important to note how imperative it is for a researcher to properly designate these groups if accurate depictions are to be made. 

One way to clarify the importance of accurate and truly representative grouping is to consider the ancient Chinese text, The Art of War.  Originally, this text was comprised of a relatively small number of comments made by the famous Chinese general Sun Tzu, but as the text grew in importance throughout the ages because of its apt teachings on military strategy and leadership, other of China’s well known and successful generals began to add their commentary as well.  Today, copies of The Art of War come with these other generals comments layered in on the pages with Sun Tzu’s original commentary, and the result is an amalgamation of China’s greatest military strategists ideas as they follow and evolve from the original authors.  If one were to graph these generals and to measure their successes in the chart, then it would likely appear that SunTzu’s philosophies were so astute that they have inspired nothing but success.  This is a problem.

The problem is not that so many generals have followed in Sun Tzu’s footsteps by adding to and amending his writings, but since the text has been so influential in terms of its acceptance as a military guidebook (taught in military academies around the world) there have been many unsuccessful generals that have adopted and enacted Sun Tzu’s teachings as well.  If the graph only consisted of those generals included in the work, then a false depiction of the history of Sun Tzu’s strategies would be the product.  This is not so different from any other type of statistical fallacies that are so often encountered in charts of this kind, and it is crucial that those interested in literary research of this kind understand both their subject and statistics.  In the end, I think that Moretti makes good use of his charts and data sets, but if considered closely enough certain fallacies can be found in them as well.  For instance, in his graph of literary genres, in which he tries to chart the emergence and decline of genre types, I’m not so sure that he allows for trend cycles that include the potential for genre comebacks or tapering interests.  This is a definite issue with this type of research, but as the digital humanities progress and technology becomes better at allowing for this kind of confounding variable, accuracy should improve.