Data Visualization Using Voyant Tools to Analyze George R. R. Martin’s Game of Thrones novels
First publishing in 1996, the series A Song of Ice and Fire debuted with the first installment, A Game of Thrones. The initial text was met with acclaim and precisely prepared readers anonymously by the review site Kirkus in 2010 to “be prepared for a lengthy series with an indefinitely deferred conclusion.” Readers today who enjoy the series await the final two installments from Martin, although the television series based on the books has already concluded, offering a potential ending still unconfirmed.
For this data visualization project, this corpus of five novels spanning over one million total words with a readability index of 6.817 and averaging 12 words per sentence. With novels replete with characters speaking to one another, Martin’s series is no exception as dialog causes the word “said” to be the most frequently used word.
While the predictability that a novel would have “said” as the most frequent word, subtleties can be identified which correspond to plot elements in the series. Initially, much dialog occurs as backstory and character development is driven by conversation. Martin creates the low point just over the halfway point of the series by being more descriptive with character quotations, shifting from the more frequent “said” toward “observed” or “commanded.” At times he simply leaves off quotation description to convey a cold tone to the reader as much of this section dialog involves speaking from royalty to servants.
Of the other most frequent words, titular honorifics like “Lord” and “Ser” dominate the pages. “Man” rounds out the top four and offers an important insight into Martin’s story development as a male-dominated, male-driven story. With battlefields and palaces full of men vying for power, the series experiences an intriguing lull from these top four words in the middle when one of the books turns largely from the male-dominated storyline toward a section where commoners interact with other commoners, female characters escape the palace escapades while in hiding, or when female characters are isolated and imprisoned.
Of the named characters, the top two most frequently mentioned are Jon and Tyrion, both males. Arya, a young girl, holds the third spot, with her brother and sister, Bran and Sansa, rounding out the top five. In the known world from the novels, excluding later revelations from the television series, Jon is known as an illegitimate son of Arya, Bran, and Sansa’s father. Of the 5 most mentioned characters, four of the five are from the same family.
Utilizing Voyant Tools to analyze this text has proven beneficial and insightful. Practical application of this reflective tool could examine changes in writing over time, synergy of word choice, or establishing patterns in texts that otherwise would take significant amounts of time. From a practitioner’s perspective, and that of one who applies data analytics to educational problems on a daily basis, I find Voyant’s range of tools adequate for skimming the surface of a text for relationships but compared to the power of quantitative analysis tools like Tableau, R, or even SPSS, Voyant Tools insights feels cursory at best in matters of data-driven decision-making.