As the question of discriminatory influence in results lingers, I would like to redirect thoughts to what Michael Witmore wrote in one of his recent posts. In it, he speaks about a text as being “massively addressable at different levels of scale”, such as through genre, lines, or single words or even through different modes such as a phone book dress or text tagging. I am mainly reminded of his article because the texts that I have been looking at are addressable at mainly levels of scale but the results garnered from these texts are also, themselves, massively addressable. As I noted in the last post there are an infinite variety of notes that I could make on any one of the diagrams that I can make. Some examples are:
~Berowne and Faulconbridge’s lines cluster with plays and other characters normally associated with Tragedy
~Portia and Prospero’s lines appear similar to the Histories and that Portia’s lines are most like Prospero in several tests
~From the last few diagrams, there is still no concrete presence of the Late Plays as a distinct group, especially seeing that the Tempest is still apart from the rest
~Helena groups with Vincentio, Proteus, Henry VIII
~Sir John Falstaff’s lines in Henry IV Part I cluster tragically even though he is generally associated with being a comic character
And the list could go on and on. Witmore’s discussion of addressability of a text can be expanded on the results that we inherit from texts as well. For example, should we track the changes of all items in a corpus over different tests and variations? Should one item be traced over time? Should group stability, size, and members matter and, if so, at what level of the dendrogram do we define a group? Additionally, numerous other questions arise such as should we note that the top half of the diagram of clusters in ‘revcomplete’ shows single characters while bottom half shows more pairs? And finally, to hark back to Rowe’s idea of genre in the previous post, should we be looking for a model or system that can account for all variation in diagrams like these, most, or for each diagram individually?
I suppose a logical and sensible idea to accommodate these issues would be simply to return to the text and cease to try to apply programs like Docuscope and mechanics to issues best solved by people. However, as Witmore notices, we are then again confronted with the issue of addressability such as in textual “additions” like marginalia. Witmore notes that the existence of notes, dog-ears, marginalia, references to other texts, are all testament to the flexibility of the already massively addressable textual object. But then, if we are studying such a textual object objectively or rather digitally, how would we account for such features?
Ann Blair’s article on “Reading Strategies for Coping with Information Overload” gathers strategies, as well as opinions, on the issue of too many books and text. The essay begins with Adrien Baillet’s 1685 warning against too many books present in society which is an interesting thought in itself as it provokes the query of whether or not there is a break point for grasping information and if we have hit that when trying to invoke Docuscope on these texts. It then moves into the various methods already noted above such as dog-ears and marginalia and then into some extraordinary measures. For instance, Vincent Placcius, by 1689, had created a note storage closet with a multitude of hooks labeled under topical headings on which slips of note paper were attached. This freed Placcius from the normal constraints of a bound journal and allowed him the autonomy of reorganization without reduplication. Seen through the lens of Placcius’s system, pieces of Romeo and Juliet could be “filed” under the same topic of ‘young love’ as Othello or Troilus and Cressida. Does this then suggest that in our own analysis we should compartmentalize pieces rather than wholes, in terms of the plays? I fear that in all of the meanderings that have occurred in my posts and thoughts that the true objective of what is being searched for has been lost. But in a sense, I am simultaneously having a hard time finding a stable ground to start from. In a realm of textuality such as physical instances which Blair is concerned and or the intangible world of digital humanities, what can we consider an “anchor” point to be? Or, in other words, what is a truly determinate point of reference in an infinitely re-scalable diagram, text, genre, or literature on the whole?
In sum, we are not left with a specific issue of textual interpretation or decisions between instances of nested textual objects but rather the larger problem of how to address too much information, textual, graphical, and combinatory, and still make sense of it.