Comparing 3 Digital Tools

After working with Voyant, and Palladio, I feel like I have a better understanding of the kind of data that each tool works best with, some of the weaknesses of each tool, and how the tools can be used together. Voyant is meant to be used primarily with textual data. At the bare minimum, requires geographical coordinates. Palladio works to visualize relationships between different elements within a dataset or datasets and emphasizes the interconnectedness of those elements.

Shortly after completing the Voyant module, I learned about a research project involving the textual analysis of 16 of Agatha Christie’s mystery novels. Dr. Ian Lancashire used two different pieces of software (Concordance and Text Analysis Computing Tools) to analyze Christie’s writing and make observations about the author’s vocabulary size and how it changed throughout her career.  The results of the analysis led Lancashire and other researchers to theorize that Christie suffered from undiagnosed dementia at the tail end of her writing career. This example showcases the importance of textual analysis tools like Voyant; the ability to organize text in new and different ways allows researchers to view data from new perspectives and generate new questions and theories. A word of warning though – users of Voyant (and really any digital humanities tool) should still question any new patterns brought to light by the software.  For example, a quick analysis of Franz Kafka’s book The Metamorphosis in Voyant displays words like “father”, “sister”, “mother”, and “family” as some of the top words in the work. It wouldn’t be a crazy jump for someone who has never read the book to conclude based on this data alone that Kafka’s work is a touching tale about a man’s family. This scenario hits home that context matters, which is why Voyant has included a “Context” view.

Words that appear most frequently in Franz Kafka's "The Metamorphosis", not including stop words. Via Voyant and Project Gutenberg.
Words that appear most frequently in Franz Kafka’s “The Metamorphosis”, not including stop words. Via Voyant and Project Gutenberg.

The overlap between Voyant and (and other GIS analysis tools) is clear: as long as a dataset contains text and geographical locations, both tools can be utilized. Take for example, the Authorial London project which maps locations in London referenced in literature. If a user wanted to analyze literary works by Charles Dickens, they could input his works into Voyant and analyze the texts that way. But GIS tools can provide a different perspective on the same set of information. Using Authorial London, a researcher can easily view a map of the locations Dickens references in his works, including the overlap in places he had personal connections to. The use of both textual analysis tools and GIS tools paints a fuller picture for researchers and provides new perspectives that previously would have been difficult to achieve. In my opinion, is superior to the software used in Authorial London because of the flexibility it provides its users, especially with the easy-to-manipulate filters and layers functions.

Palladio is unique in that, in addition to offering graphical visualizations of networks, it also offers a map visualization of the same dataset.  The map function of Palladio doesn’t seem to be its strongest feature, so it makes sense that if a researcher wanted to examine data from a network perspective as well as a geographical perspective, that researcher would utilize both Palladio and For me, Palladio and the module about network analysis was the most challenging to wrap my mind around. It’s easier for me to see the value in network analysis when examining relationships between people and another element (other people, locations, events, topics, etc.). A great example of this is the Digital Yoknapatawpha project and the  visualizations included there analyzing the networks between characters in William Faulkner’s works and other elements in the author’s novels (locations, other characters, events). Examining relationships between non-individuals (as we did with the locations and interview topics in the WPA Alabama Slave Narratives exercise) muddies the waters for me a little bit, but perhaps things will become more clear with further exposure and practice. I was also very aware of some ease-of-use shortfalls concerning Palladio. Coming from which made it easy to color-code points on the visualization, options to do so were severely limited in Palladio.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.