Data Mining and Quantifying Literature
Alex Forbes
ENGL 510
Malinda White
September 29th
Drucker emphasizes the importance of quantifying literature, and how it can help readers to better understand what they are reading, "the act of quantifying literature, while it may seem reductive, opens up new avenues for understanding texts" (112). Digital tools like Voyant help to reveal patterns and connections that you may not be able to interpret through traditional reading methods. Obtaining quantitative data from literature can be both enlightening as well as controversial. For instance, while tools like Voyant offer a way to visualize and analyze vast amounts of text, allowing us to identify trends and themes that might not be immediately apparent, Voyant also gives us a narrow view of the text, failing to provide context that may be essential to understanding the quantitative data. For instance, through text mining, you can discern the frequency of specific words or phrases thus offering insights into repeated themes or shifts in language over time. Drucker notes, "data mining can reveal undercurrents within a body of work that may not surface in traditional analyses” thus proving that it is important to have both data mining and quantitative literature as well as traditional text, because neither are useful when used alone.
The simplification of complex literary works into mere data points highlights the importance as well as debate over interpretation and meaning. The risk with analyzing text solely through the use of quantitative data is that the information could be oversimplifying the nature of the original text. Drucker cautions against viewing data as valuable without the context provided by critical interpretation, "data without context can lead to misinterpretations, underscoring the need for a synthesis of quantitative and qualitative approaches." This statement emphasizes the necessity to include traditional literary methods with digital analysis, since each approach works to further one’s understanding of the text.
Applying data mining to traditional literary research can increase your understanding of authors, genres, and historical contexts. For example, by using Voyant, you can track how a particular author’s use of language evolves throughout their work and how thematic elements align with historical events. When using Voyant, I found the tool’s visualizations, such as word clouds and trend graphs to be useful in better understanding the meaning of the story as a whole, however it wasn’t something that I realized my first time using it. It took me a while to navigate the software as well as to employ that knowledge to better comprehend what the data is showing me. As I analyzed the data more, I began to notice certain trends and associations between the frequency of certain words being used which helped me to better apprehend the story and see what the bigger picture might be. It’s interesting seeing how there is a bit of math and science that is involved in writing a story.
By utilizing tools like Voyant alongside conventional critical practices, you can uncover a layered understanding of texts that engages with the patterns that data mining reveals. As Drucker hints, it’s possible that the future of literary studies may lie in this combination of quantitative analysis and qualitative inquiry. Since analysis helps illuminate inquiry, thus enhancing one’s understanding of the literary work as a whole.
From Gabby: Responding to Alex Forbes’ Blog Post on ‘Data Mining and Quantifying Literature,’ I really enjoyed the context on both the pros and cons of quantitative data through literature and providing examples through Voyant on how those patterns can be both providing yet damaging because of the lack of context we are intaking within the actual text itself. Whereas on the other hand it's beneficial because it gives quick, accessible information that is needed about the text. Overall a very interesting take on this chapter. I love the quote, “The risk with analyzing text solely through the use of quantitative data is that the information could be oversimplifying the nature of the original text.” Because I think that accurately describes the understanding of quantitative data within text and the consequences it could have even though it was made as a good, helpful thing.
ReplyDeleteI like how you elaborate on how quantitative methods can reveal patterns and themes that might otherwise go unnoticed. Voyant seems to be a great tool in this way, but you have to draw a narrative out of it. For example, my assigned story, Afterward by Edith Wharton has a very confusing word frequency chart at first glance. The most common words where long, know, and house. When reviewing that data, I had little to no clue how to pull meaning out of these words based on their frequency. To help myself, I focused on one at a time. This is where I found that house was talked about far less frequently towards the end of the story, implying that the house that they moved into, grew less important at the end of the story. This falls in line with the word "know" which I noticed was paired more they a few times with the idea of actually "not knowing". Their house was haunted, so the actually house became less important, and the idea of "not knowing" become more. A narrative element brought forth by Voyants usage of qualitative data analysis.
ReplyDelete